The Sago MineDisaster

A preliminary report to Governor Joe Manchin III

J. Davitt McAteerand associates

July • 2006

The Sago Mine DisasterA preliminary report to Governor Joe Manchin III

J. Davitt McAteer

and associates:

Thomas N. Bethell

Celeste Monforton

Joseph W. Pavlovich

Deborah Roberts

Beth Spence

JULY • 2006

Buckhannon, West Virginia

This report, as well as additional related information, is available at:

www.wvgov.org and www.wju.edu

Front cover photo:

Memorial ribbons on the Sago Mine security fence, January 2006

Jeff Swensen / Getty Images

Back cover photo:

Mourners praying at the funeral of Jerry Groves, January 2006

Haraz Ghanbari / AP Images

On January 12, 2006, West Virginia Senate President Earl Ray

Tomblin (D-Chapmanville) and House Speaker Bob Kiss (D-Raleigh)

appointed Senators Don Caruth (R-Mercer), Jeff Kessler

(D-Marshall) and Shirley Love (D-Fayette) and Delegates Mike

Caputo (D-Marion), Eustace Frederick (D-Mercer) and Bill Hamilton

(R-Upshur) to conduct an inquiry into the Sago disaster. At the

request of Governor Joe Manchin III, their inquiry was conducted

jointly with our investigation. These legislators have worked

diligently with us in seeking answers to this West Virginia tragedy.

1

Contents

Letter of transmittal ..........................................................2

Dedication........................................................................4

1. Executive Summary .......................................................7

2. Recommendations ....................................................... 17

3. Synopsis .................................................................... 23

4. What caused the disaster? ............................................ 35

5. Why did the seals fail? ................................................. 41

6. Did the miners’ SCSRs fail? .......................................... 49

7. Why did the mine rescue effort take so long? .................. 57

8. What happened to the three shots? ............................... 83

Appendix ....................................................................... 92

Acknowledgements ......................................................... 96

2

Letter of transmittal

Governor Joe Manchin IIIState of West Virginia

1900 Kanawha Boulevard East

Charleston, WV 25305

Dear Governor Manchin:

I have the honor to submit to you this preliminary report on the Sago Mine disaster of

January 2-4, 2006.

On January 9, 2006, when you appointed me as your special advisor on the disaster,

you directed me to investigate and report on the causes; the rescue and recovery

operation, including the miscommunication that caused the Sago miners’ families such

needless heartache; and the steps needed to improve mine safety and mine communi-

cations in West Virginia and across the nation.

This report seeks to address these concerns. My hope is that it honors the pledge you

made to the victims’ families when you stated: “These twelve lives will not be lost in

vain. It is my goal to see that West Virginia is home to the safest mining operations in

the nation.”

I believe we can fulfill this pledge and accomplish this goal. But the task will require

dedication, innovation, and vigilance on the part of all in the mining community.

For me, and for my associates in this investigation, it has been an inspiration to meet

and work with the families of the lost miners. In their grief they have become tireless

advocates for change. They have our deep respect and support.

The miners’ families insisted that their loved ones be honored and remembered. They

demanded that the disaster be investigated thoroughly and in the clear light of day.

They were your strongest allies as you and the legislature moved with unprecedented

speed to enact new mine safety legislation on January 23; they were full partners in

the public hearings we conducted on May 2-4; and they became a vital presence in

Washington, working with West Virginia’s congressional delegation to pass the new

federal legislation enacted by Congress in June. As we look to the future, miners’

families have the most at stake and are the main reason why I can feel confident that

real change is coming to the coal industry.

This is a preliminary report primarily because we do not yet have all the answers

about the disaster’s proximate cause. But in your initial charge to me you directed that

if, during the course of the investigation, we were able to reach conclusions about

improvements to mine safety and health which could be implemented along the way,

we should make those conclusions available. We are doing so in this report, which

includes numerous recommendations to improve mine safety and mine rescue.

3

We offer these recommendations in full awareness that more remains to be done. But

that should not be a rationale for delay. There are technologies and equipment avail-

able today that could be making our coal mines significantly safer tomorrow — for

example by improving communication and the ability to locate miners, and, when

escape is not possible, improving miners’ chances of surviving an explosion or fire by

providing refuge chambers.

It has been suggested that some of this equipment is “unreliable,” by which is meant

that it may not work in all locations at all times. That criticism is misplaced.

Underground coal mines are both similar and unique. The roughly 150 currently active

underground coal mines in West Virginia are similar in purpose and function. They are

different, even unique, in the geological, engineering, and other challenges that they

represent. The similarities suggest that innovative technologies should be broadly

applicable. The differences suggest that some technologies will need to be tweaked or

modified in order to be made reliable in specific situations.

If we insist on waiting for perfect technological answers to the challenges facing us,

we will wait forever. The unmistakable message of the Sago Mine disaster is that we

cannot afford to wait. In this report we discuss technologies and techniques which are

currently available, tested in underground mines in the United States and other mining

countries, and which would immediately improve underground miners’ chances to

escape or be rescued in an emergency. Incremental improvements should be adopted

immediately while the search for better technologies goes forward.

Because of your leadership, the leadership of the state legislature, and, at the national

level, the leadership of Senator Robert C. Byrd and the entire West Virginia congres-

sional delegation, West Virginia is poised to lead the coal mining industry of the United

States in creating a far safer workplace, one worthy of the 21st century and the min-

ers who do so much to power the nation’s economy. My hope is that this report will

help spur the kind of progress that miners have been waiting for.

The work of fulfilling our pledge to the Sago miners and their families is ongoing, and

I expect to supplement this report as necessary.

Thank you for asking me to assist you in striving to make West Virginia the nation’s

leader in coal mine health and safety. I submit this report in the hope and belief that

the Sago Mine disaster will be remembered as a turning point as well as a tragedy.

J. Davitt McAteer

Shepherdstown, West Virginia

July 19, 2006

4

We dedicate this report

to the families of the twelve good men

who lost their lives in the Sago Mine

January 2-4, 2006

Tom Anderson was ashuttle car operator on theTwo Left section of the SagoMine, moving coal from thearea of the working face to theconveyor belt. He had 10 years

in the mines. He lived in Rock Cave, UpshurCounty. Married to Lynda Hyre Anderson,he was the father of four sons, Caleb (de-ceased), Randy, Mitchell, and Thomas Isaac.He was 39.

James Bennett was ashuttle car operator with morethan 25 years of miningexperience. He lived in Volga,Barbour County. Married toLily Foster Bennett, he was the

father of a daughter, Ann Merideth, and ason, John. The eldest of the Two Left crew, he had been planning to retire in 2006. Hewas 61.

Marty Bennett was acontinuous-miner operator,running the machine that cutcoal from the face. He had29 years of mining experience.He lived in Buckhannon,

Upshur County. Married to Judy Ann LantzBennett, he was the father of a son, Russell,who also worked at the Sago Mine. Hewas 51.

Jerry Lee Groves was aroof bolter, with nearly 30years of mining experience. Helived in Cleveland, WebsterCounty. Married to Deborah A.Groves, he was the father of a

daughter, Shelly Rose. He was 56.

George Junior Hamner

was a shuttle car operator,with 28 years of miningexperience. He grew up on afarm near the site of the SagoMine and owned a small cattle

farm in Glady Fork, Upshur County. Marriedto Deborah Hamner, he was the father of adaughter, Sara Bailey. He was 54.

Terry Helms was a firebossand beltman. He had 29 yearsof mining experience. He livedin Newburg, Preston County.The father of a daughter,Amber, and a son, Nick, from

his previous marriage, he was engaged to bemarried to Virginia Moore. He was 50.

5

Jesse L. Jones was a roofbolter, with 16 years of miningexperience. He lived inPickens, Randolph County. Hewas the father of two daugh-ters, Sarah and Katelyn. His

brother, Owen, was the foreman of the OneLeft crew on the morning of January 2, andtried to reach Jesse after the explosion. Hewas 44.

David Lewis was a roofbolter, with about two years ofmining experience. He lived inThornton, Taylor County.Trained as a diesel mechanic,he went to work at Sago so

that his wife, Samantha, could stay homewith their three daughters, Kayla, Shelby, andKelsie, while she completed a master’s degreein health care administration. He was 28.

Martin Toler, Jr. was thesection foreman on Two Left,with 32 years of mining experi-ence. He lived in Flatwoods,Braxton County. Married toMary Lou Toler, he was the

father of a daughter, Courtney, and a son,Chris, who had worked with his father inanother mine. He was 51.

Fred G. Ware, Jr. was acontinuous-miner operator,with 37 years of mining experi-ence. He lived in Tallmansville,Upshur County, just across theriver from the Sago Mine. He

was the father of a daughter, Peggy Cohen,and a son, Darrell. He was 59.

Jackie Weaver was thesection electrician on Two Left,with 26 years of mining experi-ence. He lived in Philippi,Barbour County. Married toCharlotte Poe Weaver, he was

the father of a daughter, Rebecca, and a son,Justin. He was 51.

Marshall Winans wasa utility man and scoopoperator, with 10 years ofmining experience. He livedin Belington, the TalbottCommunity, Barbour County.

Married to Pamela Pharis Winans, he was thefather of three daughters, Tiffany, Mandy, andHolly. He was 50.

6

We also dedicate this report to

the sole survivor of the Two Left crew,

Randal McCloy, Jr.

with best wishes for his continuing recovery

andto the miners

who escaped the Sago Mineon the morning of January 2, 2006:

Denver Anderson*

Paul Avington*

John Boni

Pat Boni

Gary Carpenter*

Ron Grall

Randy Helmick*

Eric Hess*

Fred Jamison

Owen Jones*

Hoy Keith*

Roger Perry*

Gary Rowan*

Joe Ryan*

Chris Tenney*

Alton Wamsley*

*One Left crew

andto the miners

who courageously attempted to rescue the Two Left crew:

Jeff Toler

Vernon Hofer

Owen Jones

Al Schoonover

Dick Wilfong

7

1Executive Summary

OVERVIEW

Twenty-nine coal miners went undergroundat International Coal Group’s Sago Mine nearBuckhannon in Upshur County, West Vir-ginia, on the morning of January 2, 2006. At6:26 a.m., a methane ignition in a recentlysealed area of the mine triggered an explosionthat blew out the seals and propelled smoke,dust, debris and lethal carbon monoxide intothe working sections of the mine. One minerwas killed by the blast. Sixteen escaped.Twelve were unable to escape and retreated toawait rescue behind a curtain at the face of theTwo Left section. Mine rescuers found thetrapped miners approximately 41 hours later.By that time all but one had succumbed fromcarbon monoxide asphyxiation.

This is a report about what happened to theminers and why. It is also a report about whythe explosion was not contained; why theTwo Left crew was unable to escape; why themine rescue effort took so long; why thetrapped miners were sent no signal to tellthem that help was coming; and why, afterthey were finally found, their waiting familieswere informed that all were alive — andcelebrated that miraculous news for nearlythree hours before being told the awful truth.

This is also a report about what the Sago

disaster has taught us and what we must dowith what we have learned.

The staggering tragedy of the Sago Minedisaster — first the loss of the miners, then theappalling miscommunication to their families— shocked West Virginia and the nation.Governor Joe Manchin III pledged to thefamilies that their loved ones would not die invain. To make good on that pledge, he vowedto bring West Virginia to a position of nationaland international leadership in coal minesafety and health.

Action followed. As official investigations bythe West Virginia Office of Miners’ Health,Safety & Training (WVOMHST) and thefederal Mine Safety and Health Administra-tion (MSHA) got under way, the governorappointed a West Virginian and formerMSHA chief, Davitt McAteer, as his specialadvisor. Then, on January 23, the governorintroduced and the state legislature enacted— within the unprecedented span of a singleday — legislation requiring rapid notificationof and response to mine emergencies, storageof additional self-contained self-rescuers(SCSRs) underground, and installation ofimproved mine communication and trackingsystems in order to be able to locate andmaintain contact with miners in emergencysituations.

8

Just four days previously, on January 19, theurgent need for action had been underscoredby a fire that left two miners dead at MasseyEnergy’s Aracoma Alma Mine in LoganCounty. Since that time, five more WestVirginians have lost their lives in mine acci-dents, bringing to 19 the number of coalminers killed on the job in the state in the firstsix months of 2006 — compared to three in allof 2005. Nationwide, 34 coal miners havebeen killed on the job thus far this year,compared to 22 in all of 2005.

Governor Manchin pledged further action,announcing that public hearings would beheld on the Sago Mine disaster and thatWheeling Jesuit University would host thefirst International Mining Health and SafetySymposium in April, to help spur adoption ofimproved technologies and equipment inWest Virginia mines. WVOMHST issued anemergency rule on SCSRs, communication,and other mine safety improvements, and thegovernor appointed a six-person labor-management task force to assess the availabil-ity of the technologies covered by the newWVOMHST rule.

Meanwhile, at the national level, SenatorsRobert C. Byrd and John D. Rockefeller IVand Representatives Nick Rahall, AlanMollohan, and Shelley Moore Capito intro-duced bills to enhance coal mine safety.Hearings were held in both the Senate and theHouse, and Representative George Miller ofCalifornia convened a forum to hear testi-mony from the Sago miners’ families as wellas from family members of the 13 minerskilled in an explosion at the Jim WalterResources No. 1 Mine in Brookwood, Ala-bama, in 2001. Witnesses urged lawmakers tostrengthen oversight and enforcement of the

Federal Mine Safety & Health Act of 1977,require improvements in mine safety equip-ment and technologies, and improve minerescue response strategies and management.

The public hearings promised by GovernorManchin, originally scheduled for March,were postponed at the request of the miners’families to allow more time for state andfederal investigators to compile additionalinformation on the causes and consequencesof the Sago explosion. The internationalsymposium on mine health and safety tech-nologies went forward as scheduled and,among other accomplishments, demonstratedthat technologies available today, in areassuch as communications, tracking, andrefuges, can be put into the mines of WestVirginia on an accelerated basis to improveminers’ chances of escaping or otherwisesurviving an emergency.1

Public hearings were convened at WestVirginia Wesleyan College in Buckhannon onMay 2, 2006. Over the course of the next threedays, miners’ family members, miners’representatives, West Virginia legislators andstate and federal investigators heard detailedpresentations and statements about thedisaster. They closely questioned company,state, and federal officials about conditions atthe mine prior to the disaster, the events ofJanuary 2-4, 2006, and the actions of the minerescue command center where company,state, and federal officials shared decision-making responsibilities throughout the minerescue operation.

For the first time in the history of coal minedisaster investigations in the United States,family members participated in a publichearing on an equal basis with state and

9

Sago Mine, January 2, 2006

Two Left Working Section

▲

One Left Working Section

▲

Track

Recentlysealedarea

Omegaseals

Older SealedAreas

2 M

iles

▲

Portal

▲

▲

▲

▲

▲

58 block

▲

10

federal investigators. The 1,263-page hearingtranscript reflects both their grief and theirdetermination to honor the memory of theirloved ones. 2

As John Groves, brother of disaster victimJerry Groves, stated at the opening of thehearing: “Even though he’s gone now, he’sgoing to be helping people, because we’regoing to be, through him, working to makesure that a disaster like this never happensagain.”3

Other actions followed. On May 5, 2006, theWest Virginia Board of Coal Mine Health andSafety filed a rule to establish two state-runmine rescue teams to enhance the largelycompany-based system currently and tradi-tionally in effect. On May 12, 2006,WVOMHST Acting Director James M. Deanannounced a statewide moratorium on theuse of Omega blocks — the materials used toconstruct the seals at Sago — as alternativeexplosion-proof seals.

On May 22, 2006 — two days after an explo-sion at a Kentucky mine blew out sealsconstructed of Omega blocks and killed fiveminers — MSHA announced a nationwidemoratorium on the use of alternative sealspending a reassessment of their ability towithstand explosions. On May 29, 2006, theWest Virginia Mine Safety Technology TaskForce issued its report. Among its otherrecommendations, the task force called foradditional provision of SCSRs and improvedtraining; emergency refuge shelters; and anAugust 2007 deadline for mine operators tosubmit plans for improved communicationand tracking of miners underground.

On June 15, 2006, at a White House ceremony

attended by several of the Sago victims’families as well as Governor Manchin andWest Virginia’s congressional delegation,President George W. Bush signed into law theMine Improvement and New EmergencyResponse (MINER) Act. Among other provi-sions, the new federal law requires under-ground miners to be supplied with two hoursof emergency oxygen at their work sites, aswell as additional oxygen cached inescapeways; requires mine rescue plans to beregularly updated and mine rescue teams tobe within an hour’s travel time from themine; and allows MSHA to seek injunctionsto halt production at mines whose ownersignore outstanding fines.

Meanwhile, investigative work on the SagoMine disaster continued, and a long-awaitedinterview with the sole survivor, RandalMcCloy, Jr., eventually made possible by hisremarkable recovery, was held on June 19,2006.4 The release of this report, originallyscheduled for July 1, 2006, was briefly post-poned in order to allow more time to attemptto determine the proximate cause or causes ofthe Sago explosion and to follow up onconcerns raised in the interview with Mr.McCloy, primarily about the failure of four ofthe SCSRs relied upon by the trapped minersand the failure of the command center to sendsignals to them to let them know that rescuerswere on the way.

This report is being released now with theproximate cause of the Sago explosion stillunsettled. There appears to be little doubt thatan explosive mixture of methane and airignited in the recently sealed area of the mineinby the Two Left section at 6:26 a.m. onJanuary 2, and the belief is widespread that apowerful lightning strike was involved. As of

11

this writing, however, the means by whichlightning may have entered the sealed area atthat moment has not been determined.Although various theories have been offeredby International Coal Group consultants andothers, definitive evidence has not yet beenfound or confirmed by state or federal investi-gators. Consequently this report does notpurport to identify the proximate cause of themethane ignition — and, therefore, untildefinitive evidence becomes available, thisreport must be considered preliminary. Oncethe proximate cause has been identified,further work will be required in order to beable to make specific recommendations aboutthe steps necessary to reliably protect under-ground mines and miners against a repetitionof the Sago Mine disaster.

THE DISASTER: CAUSESAND CONSEQUENCES

It is important to make a clear distinctionbetween what caused the explosion and whatcaused the disaster.

There would have been no disaster if theexplosion had been contained by the suppos-edly explosion-proof seals. If that had hap-pened, it is possible that all of the miners inthe Sago Mine at the time of the explosioncould have escaped safely and successfully.

There might have been no disaster if theminers on Two Left had been able to make allof their SCSRs work. Randal McCloy saysthat they could not. That left them little choicebut to await rescue, which then came too latefor all but one of them.

There would have been no disaster if it hadbeen possible for rescuers to communicate

with the miners on Two Left. They could havebeen told that there was respirable air at themouth of the section, not far from where theywere. But the mine’s phone system — anoutmoded, vulnerable system typical ofunderground mines — was knocked out bythe explosion.

There would have been no disaster if minerescue teams had been sent into the minesooner and allowed to advance to Two Leftwithout being required to systematicallyassess post-explosion conditions throughoutthe mine on the way in. But there were delaysin organizing the rescue, and the absence of agas chromatograph for much of the first daymeant that gases indicative of the presence orabsence of a fire could not be accuratelymeasured. The command center chose acautious, time-consuming approach geared toprotecting rather than risking the lives of themine rescue team members. But that ap-proach, whatever its merits, greatly dimin-ished the chances of winning the race againstdeadly carbon monoxide.

There might have been no disaster if it hadbeen possible to rapidly pinpoint the locationof the trapped miners and drill a boreholedown to them. But the seismic equipmenttheoretically capable of hearing their responseto a prearranged signal was not brought tothe mine. Even if it had been brought there,it would have taken too long to set up andoperate to be of any value — but the trappedminers had never been told that. Conse-quently they had reason to hope that barri-cading themselves would work. Had theyknown the truth, they might have tried tomake their way out of the mine, even thoughthere was no certainty that they would havesucceeded.

12

There might have been no disaster if therehad been even a single stroke of good luckat Sago.

Because time is always the enemy, no minerescue can succeed without a large measure ofluck. Everything needs to go just right. AtSago, everything that could go wrong did gowrong. The explosion occurred on a federalholiday. Time was wasted trying to reachresponsible officials, determining what hadhappened, organizing a response and callingin the most experienced mine rescue teams inWest Virginia. Time was spent making edu-cated guesses about whether there was a firein the mine and whether it might trigger asecond explosion. Time was spent trying toprotect mine rescue teams against an unde-fined risk. All of that was time that the minerson Two Left did not have.

We do not find fault in this report withanyone who responded to the Sago Mineemergency. The failures we describe aboveand throughout this report were, in almostevery instance, failures not of individualhuman beings but of systems: mine safetysystems, mine emergency managementsystems, and mine rescue systems that are indesperate need of immediate help. One of theinescapable lessons of the Sago Mine disasteris that it could happen again tomorrow —and if it were to happen, the outcome nexttime might not be much different, regardlessof who happens to be in the command center.

Next time, possibly, the seals would with-stand the explosion. Possibly the SCSRswould work as expected. Possibly therewould be no failure of communications. Ifminers did find escape to be impossible,possibly a rescue command center would be

rapidly operational and ready to deploy minerescue teams within two or three hours afterthe explosion instead of six or seven. Possiblythe command center would have better datato work with and might be inclined to takemore risks. Possibly the miners would bebrought out alive.

But, as we know, none of these hopefulpossibilities came true at Sago, and theymight not come true next time. It thus be-comes a matter of great urgency to under-stand the reasons and to take corrective actionas rapidly as possible.

Accordingly, the thrust of this report isdirected toward what can be done now to getbetter safety technology, equipment, andstrategies into the mines of West Virginiawithout further delay. We know that bettertechnologies are on the horizon (they alwaysare), but we do not consider it acceptable towait for the perfection of technologies orequipment before improving the protectionavailable to miners.

FINDINGS ANDRECOMMENDATIONS

Based on the evidence available to us thus far,we do not believe that the Sago Mine disastercan be attributed to any specific actions on thepart of International Coal Group (ICG), thefederal Mine Safety and Health Administra-tion (MSHA), or the West Virginia Office ofMiners’ Health, Safety and Training(WVOMHST). This does not rule out thepossibility of further review — for example, ifit is determined (as it has not been yet) thatICG’s apparent failure to fully protect themine’s electrical and conductive infrastruc-

13

ture with the grounding and lightning arres-tors required by regulations played a directrole in providing a path for lightning to enterthe sealed area of the mine and trigger amethane ignition.

Similarly, it can be argued that a thoroughreview of the process by which MSHA andthe National Institute of Occupational Safetyand Health (NIOSH) tested and approvedOmega blocks as “explosion-proof” sealswithin the meaning of federal regulations(discussed in section 5 of this report) mayfind that the testing protocols and the criteriaupon which those protocols were based wereinherently flawed and inadequate. If so,serious questions might be raised about thepossibility of failure at the federal level.Pending such a review, however, it ispremature to make any such charges orassumptions.

By the same token, the apparent failure ofsome of the SCSRs relied upon by the trappedSago miners (discussed in section 6) raisesquestions about whether this particular unitcan be counted upon to work in a crisis — or,alternatively, whether miners are being givensufficient training on what they should expectto experience while using any rebreathingdevice. Pending a nationwide audit of SCSRs,these questions and their potential implica-tions cannot be addressed to anyone’s satis-faction. (On July 10, WVOMHST instructedall operators of underground mines in WestVirginia to begin inventorying the estimated10,000 SCSRs currently in use in the state, inorder to facilitate a quarterly functional-statusaudit,5 and as this report goes to press anationwide SCSR audit is expected.)

Although issues such as these are discussedthroughout this report and cannot be ignored,we believe that it will be more productive andprotective of miners’ safety to focus on whatshould be done immediately to address thelargely systemic failures that unquestionablycontributed to the tragic loss of life at themine. While not ruling out the possible needfor additional state or federal legislation atsome point, this report does not call for morelaws now. We find, rather, that West Virginia’sleaders and the state’s mining communityhave it within their power to implement all ofthe recommendations that we offer below.

The keys to immediate progress, in our view,are commitment, cooperation, and a willing-ness to pursue what we call ‘default options’— that is, common-sense changes that canimprove safety while we await the arrival ofbetter technologies, equipment, and strate-gies. We have faith in miners’ ingenuity, andmany of our recommendations are groundedin this approach. Four examples follow:

1. Strengthening underground

communications We do not necessarilyneed new laws or more research in order toimprove the odds that a mine phone systemwill survive an explosion. One possible‘default option,’ at least in some mines, is tobury a phone cable under the floor or trackand harden the phone handsets by placingthem in military-grade housings and/orrecessing them in the ribs of the mine. Few ifany mine explosions have disrupted the minefloor at any significant distance from thesource, and a recessed phone would stand atleast a reasonable chance of surviving thepressures of an explosion wave travelingthrough an entry.

14

2. Improving SCSR reliability Pend-ing the outcome of a national audit of SCSRscurrently in service, we can improve the oddsthat a miner’s SCSR will work when neededby making sure that miners receive thoroughtraining in the actual use of an SCSR asopposed to the current regulatory require-ment, which simply calls for being able todemonstrate the ability to open and activatean SCSR while going through the motions ofdonning it but not actually doing so. Thisgives the miner no sense at all of what it feelslike to breathe through an SCSR. Without thatexperience, a miner trying to use an SCSR forthe first time, in a frightening post-explosionor fire-threatened mine environment, mayfeel that breathing through the device feelsconstricted or even impossible. It is true thatthe SCSRs in use today need to be replaced bynext-generation designs capable of supplyingrespirable air for longer than an hour, able todock with replacement units without requir-ing removal from the user’s face, andequipped with a comfortable half-face or full-face mask that also facilitates communicationrather than the awkward, uncomfortable andspeech-inhibiting mouthpiece, nose clips, andgoggles currently supplied. But until the next-generation unit arrives in the marketplace —and unless a national audit finds serious andwidespread failures among the many thou-sands of SCSRs now in service — the ‘defaultoption’ should be to provide miners with thereal-world training they need. SCSRs that arestill functional but withdrawn from servicebecause of being past their expiration datesrepresent an ideal and immediately availableresource for training. As another part of the‘default option,’ we must require randomtesting of SCSRs currently in service byasking miners to voluntarily don them andtravel all or part way out of the mine whilewearing them.

3. Emergency shelters Right nowminers who survive an explosion such as theone at Sago have only two options: to escapeor to retreat and barricade. Getting out,obviously, is every miner’s first choice, bytraining as well as instinct, but it is not alwayspossible. Debris or a fire may block the way,or the portal may be far away, or the minermay be injured and incapable of walking agreat distance. But barricading means enter-ing a race against time, against bad odds.After an explosion or inby a fire, carbonmonoxide is almost always present in life-threatening concentrations. Only a completelyimpregnable barricade will suffice. Even ifbuilding one is feasible, the supply of respi-rable air behind it will be finite and will bedrawn down and contaminated as it isbreathed and rebreathed by the minersrelying on it. Yet, as Sago shows so clearly andtragically, barricaded miners cannot count onbeing rescued within the limited time framewhen life is still sustainable. An obvioussolution is to require emergency refugechambers or comparable shelters in under-ground mines. But — some will say — thatcould take years, until thorough testing ofdesigns and structures has been completedand state and federal regulations have beenpromulgated. Possibly so. A viable ‘defaultoption’ in the meantime, at least in somemines, may be to use mine engineering andmachinery already in place to carve a rescuechamber into the coal seam. The next stepwould be to double-reinforce access to thechamber by building two concrete-blockstoppings with offset doors, and then tosupply the chamber with air and a phone line(through boreholes), water, food such asmilitary meals-ready-to-eat (MREs), blankets,cots, first-aid supplies, and a chemical toilet.Miners should be able to survive in such a

15

shelter for days, and if no other exit option isavailable, a borehole large enough for a rescuecapsule can be drilled down to the shelter. Isthis just someone’s vague notion of some-thing that might or might not work? No. Arefuge chamber of this type and design wassuccessfully constructed in an UpshurCounty, West Virginia underground minenearly 30 years ago, as described elsewhere inthis report. Fortunately it never had to beused. The kind of ‘default option’ that wasavailable to Upshur Coals then could beavailable to other West Virginia mine opera-tors now. This is not to suggest that a one-size-fits-all approach would work, or thatrefuges can be constructed without dueconsideration of all the engineering andconstruction challenges that may be involved,but only to suggest that miners’ ingenuity canbe brought to bear on this topic as on so manyothers.

4. Emergency communications and

tracking It seems likely that in the nearfuture we will be able to install and rely onrobust, redundant two-way wireless commu-nication systems capable of surviving explo-sions. But most such systems are now at thetesting stage, and the process of actuallygetting them approved and into the minescould take longer than expected. In themeantime, there is a need to be able to locateminers and send messages to them in anemergency. A hardened mine phone system,as described above, may be one ‘defaultoption,’ but it would not be wireless or able tolocate miners missing in an emergency. AnAustralian manufacturer currently offers awireless system capable of transmitting textmessages to miners via a digital display onthe top of the miner’s cap lamp battery. Theunit flashes the miner’s cap lamp to alert him

or her to an incoming message. The samemanufacturer offers a tracker unit, no bulkierthan a pager or small cellphone, that can becarried by a miner (or mounted on mobileequipment) and that will tell the mine dis-patcher where the miner (or equipment) islocated. If the underground antenna for eithersystem is knocked out, it can be replaced byan emergency antenna laid out on the surfaceabove the mine. This ‘default option’ isavailable now and in widespread daily use asa productivity-enhancing tool — in Australia.A few mine operators in the United States arealso using it successfully, and it has savedlives in this country — notably in the success-ful evacuation of 45 miners in 45 minutesduring an underground fire at the WillowCreek mine in Utah in 1998.6 But many mineoperators are reluctant to invest in such asystem, because it may not work flawlessly inevery part of the mine at all times, because itcan only transmit one way, and because“something better” (wireless two-way voicecommunication) is on the horizon. Yes, buthorizons can be deceptively distant. In themeantime, installing an interim systemshould be considered a ‘default option’ forsafety-conscious West Virginia coal operators.

As these examples attest, significant improve-ments in mine safety may not have to wait forthe perfection of new technologies. They dorequire financial outlays, but the costs pale incomparison to the costs of losing miners indisasters. In any case, it simply makes goodsense to explore creative ‘default options’while demand for coal is surging and pricesare near or at all-time highs, as is the casetoday.

If cost is not an insurmountable barrier toadoption of ‘default options’ and the kinds of

16

safety improvements summarized below,then what stands in the way? The answer, itwould seem, is inertia.

For a variety of reasons, some in the coalmining industry have never seen fit to putsafety on a par with production. That mustchange — and, we believe, is changing. Allthat is needed now is the will to accelerate themomentum for change that began building inthe dark hours after the outcome of the Sagodisaster, when the governor of West Virginiapledged that the miners lost in that disasterwould not be forgotten in the way that somany thousands of miners lost in the pasthave been forgotten. We wholeheartedlysupport that pledge and seek to help fulfill it.

The shock waves from the Sago Mine disasterare still being felt —most profoundly among

the grief-stricken families still struggling tounderstand how and why such a thing couldhave happened, but also among those in themining community who know in their heartsthat we can and must do much more toensure that every miner who goes to a job inan underground mine in West Virginia cancount on coming home alive and well at theend of the shift.

Underground mines are demanding andunforgiving environments. Mine safety isabout improving the odds that the dailychallenges of mining as well as emergencieswill be met and mastered. We offer therecommendations that follow — and thisentire report — in that spirit.

SOURCES

1A summary of the Symposium is available at www.wju.edu/sago/def.asp2The transcript of the May 2-4, 2006, public hearing may be accessed atwww.wvminesafety.org/PDFs/SAGOHearing-CertifiedDay1.pdfwww.wvminesafety.org/PDFs/SAGOHearing-CertifiedDay2.pdfwww.wvminesafety.org/PDFs/SAGOHearing-CertifiedDay3.pdf

3May 2, 2006, public hearing, transcript, p. 78.

4A transcript of the June 19, 2006, interview with Randal McCloy, Jr. (as well as all of the interviewsconducted by state and federal investigators in the course of the Sago investigation thus far) may beaccessed at the WVOMHST website: www.wvminesafety.org/sagointerviews.htm

5More information is available from WVOMHST at:www.wvminesafety.org/PDFs/SCSRInventoryMemo.pdf

6For a description of the role of the PED system at the Willow Creek fire, see:www.msha.gov/minefire/wcreek2000/willowcreekfinal/willowcreekfinalrpt.pdf

17

2 Recommendations

1 Research, develop and adopt

emergency measures to enhanceprotection against explosions from lightningentering underground mines and sealedareas; review and enhance equipment perfor-mance specifications for mine power stations.

2 Permanently ban the use of

Omega blocks as seals, becausethe current 20-psi standard is inadequateand because they may not meet even thatstandard.

3 Require mine operators to

strengthen existing alternative

seals by preparing, within 90 days, a plan toconstruct solid concrete block or comparableseal structures in front of (outby) Omegablock seals, or take other appropriate precau-tions such as ventilating or inerting gases insealed abandoned areas, within reasonabledeadlines clearly stated in the operator’s planfor completing the task.

4 Evaluate the existing seal

standards and consider, at a minimum,upgrading to the 50 psi standard adopted inother mining countries.

5 Require refuge chambers: Mineoperators must develop a plan, by January 2,2007, to purchase or construct refuge

chambers, subject to state and federal ap-proval of the design, numbers, and locationsof such refuges, with the aim of having refugechambers installed by January 2, 2008.

6 Conduct a statewide review of

all Self-Contained Self-Rescuers

(SCSRs) currently in use to determineoperability and detect damage, and requireongoing, in-mine testing of SCSRs by minersvolunteering to don and breathe throughthem to assess performance.

7 Develop comprehensive

emergency plans. Every West Virginiamine must have a comprehensive mine emer-gency plan integrated with federal, state andoperator roles and tested periodically by thestate for effectiveness.

8 Ensure that miners have

two-way communications: Aggres-sively accelerate the testing, approval, andadoption of robust, redundant, wireless two-way communication systems in all under-ground mines.

9 Require implementation of

tracking systems via the ‘default option’of installing currently available one-wayelectronic personal emergency and trackingdevices.

18

10 Undertake a comprehensive

review of West Virginia’s mine

rescue systems, including regulations,training, equipment and coordination withWest Virginia’s Office of Homeland Securityand Emergency Management.

11 Require installation of lifelines

in all primary escapeways in undergroundmines, equipped with directional cones toguide miners to safety.

12 Expand research, development,

and manufacturing opportunities

in West Virginia to ensure that the state leadsthe nation in mine safety and health technol-ogy and equipment.

In addition to the above recommendations, we offerthe following steps to enhance prevention, improveelectrical safety, improve the safety of sealed areas,support escape in an emergency, advance mineemergency preparedness, strengthen the minerescue system, and enhance cooperation amongall mine health and safety stakeholders.

STEPS TO ENHANCEPREVENTION

Make risk assessment the top

priority in planning, developing, andmanaging mines. Safety and health protectionmust be systematically engineered into allmining activities.

Adopt a ‘default option’ strategy.

While awaiting availability of new technolo-gies, systematically strengthen the ones wehave now, such as hardening mine phonesystems (by burying or otherwise shieldingcables) to help protect against explosiveforces.

Put safety on a par with production

by expecting every mining company toemploy adequate numbers of safety andhealth personnel reporting directly to topmanagement and authorized to overruleproduction demands when inconsistent withsafety practices.

Ask miners. Involve miners in all aspectsof safety and health planning, design, man-agement and enforcement as a basic require-ment, not an afterthought.

Reinforce the state’s vital role in

miners’ health and safety by expand-ing WVOMHST’s responsibilities, capabili-ties, and funding for risk management,education, training, and research, and provid-ing competitive salaries for inspectors torecruit and retain them.

Build safety protection into all

equipment design, from improvingvisibility and proximity protection to provid-ing storage compartments for personal safetyequipment, such as deploying safety equip-ment in longwalls.

Upgrade miners’ personal protec-

tive safety and health equipment

and ‘tools of the trade,’ from brass identifica-tion tags to helmets, cap lamps, and self-rescuers, to evaluate their adequacy in a 21st-century-model West Virginia mining industry.

Urge mine operators to offer

handheld gas detectors to every minerworking underground. Miners choosing tocarry them must be provided appropriatetraining and must have the right to reportabnormal readings without fear of retaliation.

19

Maintain all safety equipment in

top condition, from SCSRs to gas detec-tion devices and communication systems.

Reject ‘business as usual’ attitudes.

When a carbon monoxide sensor alarms, theresponse should not be simply to ‘havesomeone check it out when they have time.’Safety protection must be the first order ofbusiness every day, not the last.

Modernize recordkeeping of preshiftexam records, equipment serial numbers, andsafety performance checks, using computer-based logs or templates developed by stateand federal agencies.

Develop opportunities to accelerate

safety technologies, evaluating andtransferring technologies used in otherindustries and other countries.

STEPS TO IMPROVEELECTRICAL SAFETY

Urge every mine operator to exam-

ine the mine’s electrical system, toidentify potential safety vulnerabilities.

Publicize lightning hazards:

WVOMHST and MSHA should publish, poston their websites, and disseminate the latestavailable information on protecting mines’electrical systems and infrastructure againstlightning strikes.

Improve information: Request research-ers from within the United States and abroadto present papers addressing questions oflightning and protection of sealed areas at the2nd International Mining Health & SafetySymposium in Wheeling in April 2007.

STEPS TO IMPROVE THESAFETY OF SEALED AREAS

Require mine operators to assess

and address the risk of allowing minersto work underground, or in certain areasunderground, during the critical time periodwhen methane in a sealed area is in theexplosive range (5-15%).

Consider a mandatory requirement

for a mining engineer to certify that a perma-nent seal was constructed according tospecifications.

STEPS TO SUPPORT ESCAPEIN AN EMERGENCY

Require improved SCSR training for

miners, by training miners to actually donand use breathable training-model SCSRs in adark, smoke-filled environment (usingavailable training modules and nontoxicsmoke).

Involve miners in field-testing

competing SCSR brands and encourageoperators to be guided by miners’ preferenceswhen purchasing SCSRs.

Evaluate the existing 10-year shelf

life allowed for SCSRs. Expired SCSRsshould be retained and used for training.

Require mine operators to store

sufficient numbers of SCSRs under-

ground. The ‘default option’ is the staterequirement of 3 hours per miner, increasingto 24 hours per miner if refuge chambers arenot installed in 2007.

20

Review current SCSR storage

requirements to ensure that SCSRs are notstored in environments likely to compromisethe units’ performance; reassess the adequacyand effectiveness of the currently required90-day examination.

Conduct an independent evaluation

of the current NIOSH protocol and proce-dures for examining SCSRs recovered inmine disasters.

Require SCSR manufacturers to

improve reliability by redesigning SCSRtemperature and moisture indicators toeliminate possible mis-indications of status.

Accelerate development of next-

generation SCSRs by establishing anawards competition for a prototype full-faceor half-face breathing apparatus for miners,capable of reliably providing respirable air formore than one hour and of docking withreplacement units in order to facilitate multi-hour use without interruption or removal ofthe facepiece; and urge researchers andmanufacturers to present papers on this topicat the 2nd International Mining Health &Safety Symposium in Wheeling in April 2007.

STEPS TO ADVANCE MINEEMERGENCY PREPAREDNESS

Improve mine-wide emergency

preparedness by requiring much morerigorous education and realistic training.

Establish company-wide, central-

ized safety-systems control cen-

ters: Urge each mine operator with morethan one operation to establish a centralized

center capable of monitoring safety systemson an ongoing basis and coordinatingcommunications and rescue efforts in anemergency.

Require specialized training, con-ducted by WVOMHST and MSHA, for anymanagement personnel who will be expectedto serve in a position of responsibility in amine emergency command center, andmodeled on compulsory continuing educa-tion training required in other professions.

Provide liaison to families: In anyrescue situation, the highest-ranking federal,state and company officials on site must bepersonally responsible to ensure that miners’families receive timely updates of accurateand confirmed information.

Reaffirm and strengthen the

tripartite mine rescue model throughwhich state, federal, and company officialsshare responsibilities under section 103(k) ofthe Mine Act, and develop and publicizeinformation to ensure that all parties under-stand their authority and responsibility.

Improve WVOMHST’s and MSHA’s

emergency management capabili-

ties by enhancing the importance of minerescue and making a mine rescue directorpart of top management.

Conduct mock-disaster drills to test

the emergency preparedness of stateand federal officials, mine operators and thecommunity. Drills should be conducted for arange of emergency situations, from partialmine evacuations to full-scale major minedisasters, to test the effectiveness of commu-nication systems, availability of personal

21

protective equipment, proficiency of key minepersonnel, and coordination with localresponders and government officials.Through post-drill evaluations, deficienciesin the emergency response plan should beidentified and improvements made withinspecified deadlines.

Evaluate the inventory of available

mine emergency operations equip-

ment (such as gas chromatographs, robots,portable seismic) and the ability of suchequipment to facilitate a rapid mine rescue;report results to the public. Retire unreliableor obsolescent equipment and replace withstate-of-the-art as necessary.

Continually evaluate and enhance

emergency equipment: At least annu-ally, and under actual mining conditions, testall mine emergency operations equipment(such as robots and seismic), and reportresults to the public.

Support timely gas chromatogra-

phy during a mine emergency byrequiring every underground coal mine inWest Virginia to include in its mine emer-gency preparedness plan a method to have agas chromatograph on site within two hoursafter the onset of an emergency.

Require post-emergency medical

examinations and treatment as necessaryfor any miner or rescuer exposed to toxicsubstances or other threats to life and healthin the course of a mine emergency.

STEPS TO STRENGTHEN THEMINE RESCUE SYSTEM

Establish a West Virginia-based

National Mine Rescue Committee

made up of mine rescue team members andother experts, to systematically review themine rescue system’s structure, equipment,and response strategies, and to make recom-mendations to strengthen the system, in areport to be delivered at the 2nd InternationalMining Health & Safety Symposium in 2007.

Develop a skills-ranked or tiered

scale for mine rescue teams, similar towhat is used for other emergency responseorganizations.

Enhance mine rescue training toinclude in-mine drills. Veteran mine rescueteams should be canvassed to develop a ‘wishlist’ for training and equipment that shouldbecome the basis for strengthening minerescue training and deployment.

CONSOL ENERGY’S COMMUNICATION’S CENTER

Since 1969, CONSOL Energy has operated a communications center to provide comprehensive commu-nications and monitoring services to its mining operations. The center operates on a 24/7 basis, enhanc-ing safety and productivity of CONSOL’s operations by supplementing mine-site monitoring of produc-tion, ventilation and fire prevention systems. The current communications center, located near theirMcElroy mine complex in Ohio County, West Virginia, plays an important role in preparing for andresponding to emergency situations. Well-established relationships and procedures permit the rapidmarshalling and coordination of emergency services, such as mine rescue teams, EMS, and fire andpolice departments.

22

Issue a mine rescue team member

identification card, renewable annually,to expedite travel and admission to the mineproperty.

Retire the obsolete seismic system,

and recall the hardhat stickers promisingminers that help is on the way. Promptimplementation of robust two-way communi-cations systems and tracking systems willmake the current seismic system even moreobsolete than it already is. Encourage researchon next-generation seismic technology andequipment.

Invite the Incorporated Research Institu-tions for Seismology (IRIS) and other profes-sional entities to participate in the 2nd Inter-national Mining Health & Safety Symposiumin 2007 and to present information on thefeasibility of applying seismology techniquesand equipment to mine rescue.

STEPS TO ENHANCECOOPERATION

Provide updates on a regular basis

to the Sago families and the public onprogress made to improve mine safety andthe mine rescue system in West Virginia.

Continue to host an international

mine safety and health symposium

annually or biennially to promote cooperationamong mining countries and rapid advance-ment of new technologies and techniques.

Seek participation from the interna-

tional community for a study group

to evaluate lightning risk to under-ground miners. Participants would assess theincidence and travel paths of lightning strikesnear mines and report their findings at theInternational Mine Safety & Health Sympo-sium in April 2007.

Call upon International Coal Group

(ICG) to lead the way in mine safetyand health by making its mines and otherresources available as laboratories for testingand improving technologies and trainingtechniques in cooperation with state andfederal agencies.

Examine methods to improve

accident investigation protocols andprocedures, such as appointing independentinvestigative committees and defining rightsfor victims’ families; and design a datatemplate to accompany or be incorporatedinto accident investigation reports to enhancetheir value to safety and health trainers andresearchers.

Encourage voluntary cooperation

and commitment as the preferred ap-proach to get new equipment, technology,and best practices underground with allpossible speed.

Abandon old myths and misinfor-

mation about the reliability of safety andhealth technologies and replace them withhard facts and science-based determinations.

23

3 A synopsis of the Sago Mine disaster

Note: The information in this synopsis is

derived from numerous sources, primarily

testimony provided to investigators. Transcripts

are available on the website of the West

Virginia Office of Miners’ Health, Safety and

Training at www.wvminesafety.org. The times

shown (other than the time of the explosion)

are in many cases approximate. Times, data,

etc., may be subject to further investigation

and verification.

MONDAY, JANUARY 2, 2006

3:00 am

Shortly after 3:00, firebosses Terry Helms andFred Jamison enter the Sago Mine to conductpreshift examinations of the mine’s twoworking sections, One Left and Two Left.

5:30 am

Between about 5:15 and 5:45, miners andsupervisors arrive at the mine, change intotheir work clothes, check their lights andother gear, and chat with each other in andaround the bathhouse and mine office.Meanwhile, the two firebosses report to BillChisolm, the mine dispatcher, that there areno major mechanical problems or safetyviolations on the two working sections. TerryHelms delivers his report over the mine’sphone system; Fred Jamison delivers hisreport in person after exiting the mine to re-charge the battery on his mine jeep.

6:00 am

Two crews of miners — One Left and TwoLeft, each consisting of 11 miners and asection boss (foreman) — enter the mine intwo mantrips (battery-powered rail-travelingpersonnel carriers). The Two Left mantripenters first, since it will be traveling inby(further in) to the Two Left section, past thepoint where the One Left crew will turn at atrack switch to enter the One Left section. TheOne Left crew is briefly delayed while fore-man Owen Jones, realizing that his mantrip istoo small to accommodate his crew plus threeother miners — two pumpers and a beltmaintenance man — who will be travelingpartway into the mine with him, goes backfor a larger mantrip. The One Left crew, plusthe pumpers and beltman, then follows theTwo Left crew into the mine. Fred Jamisonalso re-enters the mine to work at a conveyorbelt not far inby the portal. All told, 29 coalminers are at various locations inside the SagoMine by about 6:15 am.

6:05 am

The computer console in mine dispatcher BillChisolm’s office (a trailer separate from themine office) indicates that a carbon monoxide(CO) monitor in the One Left section is eithermalfunctioning or detecting elevated COlevels. Based on the computer display (whichflashes “malfunction” briefly before alarm-

24

ing), the dispatcher decides that the monitoris false-alarming and reportedly calls the OneLeft crew to ask the crew’s electrician to checkthe fuse when he reaches the section.

6:26 am

As a severe, fast-moving thunderstormsweeps across Upshur County, lightning boltsstrike in the immediate vicinity of the mine.The mine dispatcher, who is on the phonewith mine superintendent Jeff Toler at thatmoment, hears a loud pop on the phone andsees on his computer monitor that severalconveyor belts, which he had previouslypowered up from the computer, are shuttingdown. The computer also alerts him — bothvisually and with an audible alarm — that thebelts’ CO monitors are detecting carbonmonoxide. He reports this to Jeff Toler, whoasks him to contact the One Left and Two Leftcrews on the mine phone, to tell them tocheck the CO monitors. At first, Toler and hismaintenance supervisor, Dick Wilfong,assume that some fuses have blown, as hashappened in previous storms.

(The time shown on the computer’s clock was6:31 am. In the course of the post-disasterinvestigation, state investigators performed atime correlation with an atomic clock anddetermined that the mine’s clock was fast by 4minutes 56 seconds. According to Vaisala, anationwide lightning-monitoring service, twolightning strikes, one of which packed un-usual power, occurred within a five-mileradius of the Sago Mine portal at 6:26. Theexplosion in the mine appears to have oc-curred at virtually the exact same moment.)

6:30 am

Dispatcher Chisolm gets no response fromeither of the two crews. He calls repeatedly on

the mine phone: “Inside, inside, inside!”Within an estimated three to five minutes,One Left foreman Owen Jones calls out toreport that he and his crew have been caughtin “a big rush of air that came from inby,toward the Two Left panel,” as Jeff Tolerremembers the report. Wilfong remembersJones saying that he didn’t know whathappened. Jones recalls saying “We’ve had amine explosion in here!” and adding: “Get amine rescue team here now.” Toler tells Jonesto shepherd his crew into the escapeway.

In the mine, in the moments immediatelyfollowing the explosion, the One Left minersstruggle to see — some later describe thevisibility as less than a foot at first — and tobreathe in the dust-filled atmosphere. Sevenof the miners don their self-contained self-rescuers (SCSRs); three later describe havinghad some difficulty using them, and one ofthe One Left miners subsequently testifiesthat SCSRs should be redesigned, but fromtheir testimony to MSHA and WVOMHSTinvestigators it appears that with one possibleexception all of the SCSRs that were used bythe One Left crew functioned properly.

6:45 am

Jeff Toler, safety director Al Schoonover,maintenance chief Dick Wilfong and assistantmaintenance chief Vern Hofer enter the minein a mantrip. On the way in, in the main trackentry, they first encounter fireboss FredJamison and then pumper John Boni, bothunharmed, and instruct them to exit the mine.They then stop at a mine phone near 25 blockto call out to the dispatcher, to find outwhether the Two Left crew has been heardfrom. The dispatcher reports that nothing hasbeen heard. Meanwhile the One Left crew hasbeen making its way on foot in the escapeway

25

toward the mine portal. The miners hear thesound of Toler’s mantrip in the parallel trackentry, and a crew member, Joe Ryan, sticks hishead through a man door into the entry andcalls to Toler to tell him where they are andthat one of the crew, Hoy Keith, is havingdifficulty. Toler tells Ryan to bring the wholecrew to the mantrip and instructs Wilfongand Hofer to take the crew out.

7:00 am

The One Left crew heads out of the mine onthe mantrip. (In addition, beltman Pat Boni,after calling out to find out what has hap-pened, has been instructed to exit the mineand is walking out, unharmed, in the escape-way.) One Left foreman Owen Jones, whosebrother Jesse is on the Two Left crew, remainsin the mine, further inby, at 41 block, trying todetermine whether he can move toward TwoLeft. Jeff Toler, whose uncle, Martin Toler, Jr.,is the foreman of the Two Left crew, asksOwen Jones to come back and stay by themine phone near 30 block. Toler andSchoonover then explore inby, walking to 42block, observing damage to ventilationstoppings, and using the phone at 41 block tocall out to instruct Wilfong and Hofer to comeback with curtains and other supplies torepair ventilation controls. Toler andSchoonover then walk back to rejoin OwenJones.

Meanwhile, at about 7:00 am, the minedispatcher phones ICG assistant safetydirector John Stemple at home to tell him thatSago has a problem — that there has been asevere lightning storm, power has been lostunderground, and the One Left crew hascalled out to report trouble. Stemple, trying tounderstand the situation, asks to be patchedthrough to Jeff Toler underground. They talk

at about 7:15. Toler tells Stemple that DickWilfong recommends contacting a minerescue team and getting them started towardthe mine.

7:20 am

The mantrip carrying the One Left crewarrives at the surface. Wilfong and Hofer loadup supplies and head back underground.

Stemple begins trying to reach company,state, and federal officials. Initially he has nosuccess: Phones are not answered, or his callsgo to answering machines. In the case ofWVOMHST inspector at large Brian Mills, hegets a message that Mills’s phone has beendisconnected.

7:45 am

Stemple leaves a voicemail message forWVOMHST inspector John Collins, who livesnear the mine in Buckhannon. Collins returnsthe call at 7:50, learns of the incident, phonesBrian Mills at 7:55 to notify him (Stemple hadhad an old phone number for Mills), andheads for the mine.

8:00 am

Stemple tries to phone the Barbour CountyMine Rescue Team, which is under contractwith ICG to provide emergency services. Hediscovers that the team’s answering machineis turned off. Meanwhile, at about 8:00, ChuckDunbar, general manager of ICG’sBuckhannon division, arrives at the mine anddiscusses the situation with the mine dis-patcher.

8:10 am

Wilfong and Hofer, heading back into themine on the mantrip with supplies, meetToler, Schoonover, and Jones at 32 block. They

26

repair blown-out stoppings and proceed viamantrip to 42 block, where their CO monitorsbegin alarming. Concerned about fire and thepossibility of a second ignition, they discon-nect the batteries on the mantrip and advanceon foot inby, continuing to repair stoppings asthey go. At 49 block, Toler goes into the trackentry, and moves a phone and emergencyfirst-aid supplies from the track entry to theintake entry.

8:15 am

Meanwhile, WVOMHST inspector JohnCollins arrives at the mine. He talks withmembers of the One Left crew about whatthey experienced. When two other state mineinspectors arrive, Collins asks them to take airreadings at the return entry at the portal.They report the presence of carbon monoxide,although at low levels. Because the presenceof CO is an indication of a possible mine fireor explosion, Collins decides to issue a controlorder under Title 36, series 19, section 7.1 ofthe West Virginia Administrative Regulations.The regulation states:

Unless granted permission by the Office of

Miners’ Health, Safety & Training, no

operator may alter an accident site or an

accident-related area until completion of all

investigations pertaining to that accident,

except to the extent necessary to rescue or

recover an individual, prevent or eliminate

an imminent danger, or prevent destruction

of mining property.

Meanwhile, at about this same time, Stemplereaches a member of the Barbour Countymine rescue team. He continues trying toreach MSHA officials but encounters answer-ing machines at the District 3 office and atvarious officials’ homes. Finally, at about 8:30,he reaches MSHA inspector Jim Satterfieldat home.

8:30 am

About two hours after the explosion, at 8:32am, MSHA inspector Satterfield issues averbal 103(k) order by phone. According toStemple, the inspector tells him that no one isto enter the mine or do any work at the mine.

Section 103(k) of the Federal Mine Safety &Health Act of 1977 states:

In the event of any accident occurring in

a coal or other mine, an authorized repre-

sentative of the Secretary [of Labor], when

present, may issue such orders as he

deems appropriate to insure the safety of

any person in the coal or other mine, and

the operator of such mine shall obtain the

approval of such representative, in consul-

tation with appropriate State representa-

tives, when feasible, of any plan to recover

any person in such mine or to recover the

coal or other mine or return affected areas

of such mine to normal.

At around 8:30, Ty Coleman, ICG safetymanager for its West Virginia and Marylandoperations, arrives at the mine.

A few minutes later, at 8:37, inspector JohnCollins briefs WVOMHST director DougConaway by phone.

Meanwhile, in the mine, Toler, Schoonover,Wilfong, Hofer and Jones continue to walkinby toward the Two Left turnoff, repairingdamaged ventilation controls as they go, andpausing to let fresh air dissipate the carbonmonoxide. As they proceed they notice that itis taking longer for the CO levels to drop andthat the velocity of the fresh air coming inbehind them seems to be diminishing. Be-cause this could mean that the air is short-circuiting through damaged controls behindthem, Toler asks Jones and Hofer to go outby,retrieve curtain, and check for damaged

27

controls. Then he, Schoonover, and Wilfongproceed inby to about 57 block, where theyfind the stopping damaged and hang acurtain. They then proceed to 58 block andfind that the stopping there too is damaged.They also encounter dense, swirling smokethat does not appear to be dissipating. Theydiscuss what to do, concerned that if there is afire inby they should not be directing more airto it and possibly triggering another explo-sion. They decide, reluctantly, that the safestcourse — for the Two Left crew and forthemselves — is to withdraw from the mineand let mine rescue professionals take over.

9:30 am

Toler phones the dispatcher from 58 block,reporting that the smoke is bad and that he,Wilfong, and Schoonover are walking out ofthe mine. They walk out in the intake escape-way. There, making their way out, they findOwen Jones and Hofer waiting for them, andthe five men make their way toward thesurface together.

10:30 am

Four hours after the explosion, MSHA per-sonnel begin to arrive at the mine. The first toarrive are inspectors Jim Satterfield, PatVanover, and Ron Postalwait.

10:35 am

Toler, Schoonover, Wilfong, Jones and Hoferexit the mine. Collins and other state andfederal personnel debrief them. They alsodescribe the situation underground to ICG’sTy Coleman.

11:00 am

At about 11:00 am, more than 4 hours afterthe explosion, the first two mine rescue teams— the Barbour County teams under contract

to ICG — arrive on site and begin preparingto travel underground by noon. Meanwhile,MSHA has contacted Consol Energy, whichhas several experienced mine rescue teams,and requested that they be mobilized andsent to the mine. ICG, meanwhile, has char-tered a plane to bring the rescue team from itsViper mine in Illinois.

ICG’s John Stemple arrives at the Sago Mineat about 11:30. Sam Kitts, ICG’s vice presidentfor West Virginia and Maryland, arrives at themine at about 11:45. Shortly before noon,according to Ty Coleman’s recollection,Coleman has a command center set up and inoperation at the mine offices; he is in chargeof the rescue effort for ICG. Tim Martin, ICG’scorporate safety director, arrives shortly afternoon.

Late in the morning, ICG’s Chuck Dunbarcontacts Mike Ross, a local oil-and-gas devel-oper, to locate a drilling rig in order to drill aborehole into the mine. Ross, in turn, contactsthree drilling companies, one of which, HyreWater in Buckhannon, is requested by ICG tobring a drill to the mine site.

12:00 noon

CO levels being measured at the return entryincrease to 2,300 ppm. Elevated CO levels arealso detected in the mine offices; MSHAissues an imminent danger order, requiringthe offices to be temporarily evacuated.(However, as the evacuation is starting, COlevels decline, and the command centerremains in operation.)

1:00 pm

ICG submits a plan, approved by MSHA andWVOMHST, to continue monitoring gases atthe pit mouth. At 1:15 pm, personnel measur-

28

ing at the return record 500 ppm CO; 1.1%methane (CH4); and 19.2% oxygen.

At about this time, the command centeradopts a plan to drill a borehole from thesurface into the Two Left section, based on theassumption that the Two Left crew, if alive,are most likely inby the mouth of the section.The borehole will help determine conditionson the section and presumably can be used tocommunicate with the miners if they arefound alive. Surveying crews equipped withGPS instruments are sent out to determine asuitable location, but are hampered by unfa-vorable weather conditions, which causedifficulty and delays in getting accuratereadings from GPS geosynchronous satellitesorbiting the earth.

1:45 pm

The mine rescue team from Consol’sRobinson Run mine arrives.

2:30 pm

The mine rescue team from ICG’s Viper minearrives. Several other mine rescue teamsarrive in the course of the afternoon.

2:40 pm

Trapped miner Junior Hamner writes a note: “I’mstill OK at 2:40 p.m. I don’t know what is goingon between here and outside. We don’t hear anyattempts at drilling or rescue. The section is full ofsmoke and fumes so we can’t escape. We are allalive at this time.”

2:45 pm

Trapped miner Jackie Weaver writes: “It’s 2:45 pm— the fumes are getting terrible — everybody isstill partially OK”

3:00 pm

A gas chromatograph from Consol arrivesand is set up, facilitating bottle sampling of awide range of gases.

3:07 pm

Trapped miner James Bennett writes: “The air isbad. I don’t know how long we can last.”

4:15 pm

Gas concentrations trend down.

4:22 pm

James Bennett writes: “Time is running out for us.We have not heard anything from the outside yet.”

4:25 pm

James Bennett tries to write another note: “I loveyou...” He does not finish; it is his last note.

At some point, possibly later than this, MartinToler, Jr., writes a note: “Tell all — I see them onthe other side. Jr.” On the right side of the sheet, hewrites: “It wasn’t bad I just went to sleep.” Alongthe bottom, he writes: “I love you.”

4:45 pm

More than 10 hours after the explosion,MSHA approves a plan to send mine rescueteams into the mine for initial exploration: Tri-State Team A is to explore the first 1,000 feet,stopping every 500 feet to cross three entriestaking air readings. To eliminate conductivityhazards, the rail and belt structure will bebroken and separated by the rescuers onebreak inby the pit mouth.

29

5:00 pm

A bulldozer is delivered to the mine site andbegins to construct a 1,600-foot-long roaddown a hillside to the borehole drilling site.The bulldozer then excavates a pad 200 feetlong and 90 feet wide to accommodate thedrill. The process of surveying, road-building,excavation, and subsequent modification ofthe drill pad to facilitate precise boreholedrilling takes about 13 hours. (Under normalcircumstances, surveying, siting and drilling aborehole would take about two days.)

5:10 pm

At about 5:10, MSHA modifies the mineexploration plan to provide for the moreexperienced Consol Robinson Run team toenter instead of the Tri-State team.

5:25 pm

The Robinson Run team enters the mine inthe intake entry through the fan housing. At5:45, the team is given permission to stay inthe track entry and move inby. At 5:50, theteam is given permission not to break thetrack near the portal as originally planned, sothat the track can be used for mantrip trans-portation if needed (the track is to be brokenfurther inby).

6:00 pm

Between 6:00 and 7:00 pm — about 12 hoursafter the explosion — the Robinson Run teammoves as far as 8 block, taking readings thatindicate low levels of CO and no measurablemethane. The team is told to hold at 9 blockuntil the next team arrives. At 6:57, an accu-mulation of water is reported in the returnentry at 21 block.

7:10 pm

ICG submits a plan, approved by MSHA, to

restore power to a dewatering pump so thatthe return airway can be kept open. At 7:30,permission is given to start the pump.

8:10 pm

The pump is started.

8:20 pm

MSHA approves a plan to send two mantripsoperated by mine rescue personnel to the 17block on No. 3 belt. One mantrip will bebrought back to the surface by the exitingmine rescue team.

8:30 pm

Batteries are disconnected from a scoop andmantrip located at 17 block.

8:50 pm

Teams advance to the 32 block of No. 3 belt,where the mains reduce from 8 to 5 entries.

10:45 pm

The track is cut.

11:30 pm

On the surface, about 10 hours after survey-ing began for a suitable location to facilitatedrilling a borehole into the Two Left section,the final coordinates are determined. Workbegins to modify the drill pad to allow thedrill to set up precisely on the coordinates sothat it will drill into a mine entry and not intoa coal pillar.

TUESDAY, JANUARY 3, 2006

12:25 am

Consol’s Robinson Run B team is to advance,with Consol’s Blacksville team underground

30

for backup and the Barbour County team onthe surface for backup.

2:45 am

The mine rescue teams have advanced to 34block, about 9,400 feet from the portal, whenthey report seeing a red light and are givenpermission to approach it. They identify thelight as a CO monitoring station, previouslythought to have been de-energized. Becausethe monitoring station is a potential ignitionsource, they are instructed to withdraw fromthe mine until the system can be de-ener-gized.

2:45 am

On the surface, the drill begins drilling theborehole toward Two Left.

3:40 am

The mine rescue teams reach the surface. Inthe command center, a decision is made tokeep them outside until the borehole pen-etrates the mine.

4:00 am

The CO monitoring system’s backup powersource is successfully de-energized.

The borehole is drilled to a depth of 220 feet,about 38 feet above the roof of the mine.

5:00 am

MSHA approves a plan to load its V2 minerobot and operator and transport them to 31block. The plan is for the robot to exploreinby, followed by mine rescue personnel.

5:30 am

About 23 hours after the explosion, theborehole drill punches through into the mine,258 feet below the surface, at a point in the

Two Left section about 300 feet from the face.Air samples from the borehole indicate CO at1300 ppm (more than three times the maxi-mum safe level for a one-hour exposure),negligible methane (.04%), and 20.7% oxygen.

5:50 am

The drilling equipment is shut down and a10-minute quiet period is observed, in hopesof hearing a signal from the trapped miners.There is no signal. The drill steel is struck, tomake noise underground; there is no re-sponse.

6:30 am

About 24 hours after the explosion, minerescue teams are sent underground again, anda camera is lowered into the borehole. Itshows no blast damage in the immediatevicinity, indicating that the explosion did notoriginate inby on Two Left and suggestingthat it must have occurred in the sealed areabeyond the mouth of Two Left.

7:00 am

Between 7:00 and 8:00, mine rescue teamstravel to 31 block, No. 4 belt; the robot istransported into the mine at about 7:35;drilling of a second borehole into One Leftgets under way; and the robot reaches 32block, recording low CO levels (27 ppm).

8:50 am

The mine robot, encountering difficult terrain,is disabled.

10:20 am

About 28 hours after the explosion, thecommand center adopts a plan for the minerescue teams to proceed to 48 block of OneLeft and examine the common return inby forsix breaks. The teams are then to proceed inby

31

to the Two Left section via the intake entries.The teams will then examine the Omegablock seals. Exploration will then proceed to60 block, advancing intake air into the TwoLeft section, and will then proceed into theTwo Left section.

1:30 pm

Mine rescue teams change places: Bailey andLoveridge advance, Tri-State going outside,Barbour County providing backup on thesurface.

2:00 pm

Mine rescue teams are at 48 block in the trackentry, near the One Left section return entry.

2:15 pm

Mine rescue teams find the mantrip aban-doned by the One Left crew at 50 block,heading inby with its headlights still on.

4:30 pm

Mine rescue teams, using curtains to restoreventilation in damaged areas, establish a freshair base at 57 block, No. 4 entry.

5:15 pm

MSHA mine rescuer Ron Hixson finds aminer’s body in the track entry between 57and 58 block. The victim is later identified asTerry Helms.

6:30 pm

36 hours after the explosion, mine rescueteams find that the Omega block seals beyondTwo Left have been almost completely de-stroyed, with the pulverized seals blown in anoutby direction, indicating that the explosionoccurred in the sealed area. (The destructionis so complete that mine rescue team mem-bers at first walk past the entry into what hadbeen the sealed area without realizing it.)

7:15 pm

Mine rescue teams begin exploring the TwoLeft section.

7:45 pm

Mine rescue teams find the Two Left mantripat 10 block. Moving inby in the track entry,the teams find widespread damage tostoppings.

8:00 pm

Shortly after 8:00 pm, mine rescue teamsexploring the Two Left intake escapeway findevidence of 12 SCSRs having been opened,and see footprints heading outby toward theseals.

9:00 pm

Mine rescue teams continue to advance intothe Two Left section.

11:45 pm

Mine rescue teams have advanced 1,900 feetinto the Two Left section. As rescuers advanceto the face of the No. 3 entry, Jim Klug, cap-tain of Consol’s McElroy team, reports hear-ing sounds, and moments later he and otherrescuers find the 12 trapped miners behind acurtain. The first report back to the commandcenter, relayed by voice, walkie-talkie, minerescue lifeline and mine phone, is that all 12miners are alive. The command center eruptsin celebration and then organizes to send instretchers, blankets, water, SCSRs, first aidsupplies and medical personnel. Wordquickly reaches the miners’ families waitingnearby in the Sago Baptist Church, and theycelebrate the miraculous news: 41 hours afterthe explosion, their loved ones have beenfound alive.

32

WEDNESDAY, JANUARY 4, 2006

Midnight

Mine rescue team members, several of whomare certified emergency medical technicians(EMTs), determine that there is only onesurvivor. They place him on a stretcher, beginusing a succession of SCSRs to give him air,and prepare to carry him back to their freshair base and then to the nearest mantrip —nearly 3,000 feet in all, through difficultterrain. One of the mine rescuers, aware of themiscommunication, uses a walkie-talkie to tryto radio back to the fresh air base, but cannotget through.

12:15 am

At 12:18 am, the command center receives amessage from the mine rescue teams that theyare leaving the Two Left section, and that theyhave one coming out on a stretcher and 11“items.” The command center, concernedabout communication leaks to the media, hadearlier adopted a code by which the minerescue teams would refer to bodies as“items,” and the teams were supplied anidentification sheet with a number for eachmissing miner, to be used instead of names inradio and phone communications. ICGcorporate safety director Tim Martin, remem-bering the code, asks for confirmation. Thesame answer comes back from underground.There is disagreement within the commandcenter about its meaning. Martin tells thecommand center to drop the code and askabout survivors. At about 12:23 am therescuers are asked: “Do you have 11 survivorsand one on a stretcher?” The answer comesback: “No, it’s the other way around.” Thecommand center is staggered by the news.

12:30 am

Mine rescue teams continue bringing the solesurvivor out of the mine, reaching the surfaceat about 1:00 am. With his identity not yetdetermined, the survivor is rushed by ambu-lance to St. Joseph Hospital in Buckhannon.Meanwhile, all mine rescue teams have beenordered out of the mine.

1:45 am

At about 1:45 am, with the identity of thesurvivor still not yet known, ICG officialsattempt to send a message to the families atthe church via the state police, asking anofficer to relay word to the clergy to the effectthat the command center has conflictingreports on survivors and that the initialreports may have been too optimistic. Themessage is reportedly relayed but does notreach the families.

2:00 am

Shortly before 2:00 am, ICG’s Viper minerescue team, which has 5 EMTs equippedwith stethoscopes, is rushed back into themine to confirm that the remaining Two Leftminers are deceased.

2:30 am

Mine rescuers reach the face of No. 3 entry, re-assess the victims, and report back that all 11are deceased. Meanwhile, the commandcenter hears from the hospital that the solesurvivor is Randal McCloy, Jr.

A few minutes later, ICG president and CEOBennett Hatfield, vice president Sam Kitts,and others make their way to the church andbreak the news to the families. Facing anoutpouring of confusion, disbelief, grief, andoutrage, the ICG officials are escorted out of

33

the church by state troopers and return to thecommand center.

5:00 am

At around 5:00 am, nearly 48 hours after theexplosion, mine rescue teams carry out thetask of identifying the victims, placing themin body bags, and using mine equipment (ascoop) to bring them out because of thedifficulty of carrying them all out on stretch-ers and the risk that to do so could exhaustor otherwise endanger the rescuers. The

process of bringing the victims out of themine continues for another 5 hours.

10:00 am

All victims are removed from the mine byabout 10:00 am — nearly 52 hours after theexplosion.

After the victims have been taken away inambulances, the rescuers gather in a circle atthe portal and say a prayer for the lost miners.

34

35

Shortly before dawn on the morning ofMonday, January 2, 2006, a severe thunder-storm swept across Upshur County, WestVirginia. At 6:26 a.m., multiple lightningstrikes, including one of much greater thannormal power, were recorded within a five-mile radius of the portal of the Sago Mine,located about five miles south ofBuckhannon, the county seat. At virtually theexact same instant, methane ignited in arecently sealed area of the mine just beyondthe Two Left working section, blowing out theseals and filling the mouth of the section withsmoke, debris, fumes, and lethal levels ofcarbon monoxide.

Twenty-nine miners were underground atSago when the explosion occurred. Thirteenwere in close proximity to the blast. As wassubsequently learned, one miner was killedby the force of the explosion itself. Twelveothers, after attempting to make their way outof the Two Left section, turned back andretreated behind a curtain hung diagonallyacross the No. 3 entry at the working face ofthe section in an attempt to barricade them-selves against the smoke and carbon monox-ide created by the explosion. There theyawaited rescue — which, tragically, wouldcome too late for all but one of them.

4 What caused the disaster?

Meanwhile the other 16 miners, who weremore distant from the ignition point and notseriously harmed by the explosion, were ableto exit the mine safely with the aid of otherminers, including the mine superintendent,who rushed into the mine and who thenmade a courageous but unsuccessful effort toreach the miners trapped in the Two Leftsection.

At this writing, the proximate cause of themethane ignition has not been determinedwith finality. Among the Sago miners whoescaped (as reflected in their testimony toinvestigators), the belief is widespread thatlightning somehow triggered the methaneignition. But the question of precisely howlightning may have entered the recentlysealed area of the mine has not yet beenresolved.

Multiple lightning strikes — at least two andperhaps three — are known to have occurredin the vicinity of the mine. The most powerfulof the strikes registered at 101 kilo amps (kA),roughly three times the average peak currentof lightning, and occurred at 6:26:35 a.m. atcoordinate N.38.926 – W.80.233, approxi-mately two miles from the sealed area of themine but quite close to the 12 KV power linesowned by International Coal Group and

36

leading from the Allegheny Power substationat French Creek to the mine preparation plant,then branching off with a separate split to themine offices and portal, where it suppliedpower underground.

Two other lightning strikes were recorded at6:26:35 a.m., closer to the mine power line,preparation plant and portal but of normallightning strength of 35 kA and 38 kA. Thesemay actually have been a single strike, subjectto differing interpretations as to location.1

At the same instant — 6:26:35 a.m.2 — thecarbon monoxide (CO) monitor alongside themine conveyor belt at block 57 spiked at 51parts per million (ppm).3

These simultaneous events would seem topoint to lightning as the probable cause of themethane ignition and subsequent explosion.However, investigators are still attempting todetermine how lightning may have enteredthe sealed area.

One possibility is that lightning followed apath (still undetermined) from the surfacedownward into the sealed area, possiblyinvolving a submerged pump, partiallysubmerged pump cable, and wire roof meshin finding its way to the site of the ignition. Ade-energized and submerged pump was leftin the area when the area was sealed, report-edly because crews were unable to remove it.The pump was reportedly connected to alength of cable which reached to near theignition point.4 At this writing, the pump hasbeen removed from the mine for furtherexamination.

Another possibility is that lightning followeda path (also still undetermined) to the mine

portal, then traveling through the electricsystem, phone system, track, belt structure,and/or roof mesh, and thereby finding someway to and past the seals and into the sealedarea to the point of the explosion.5

Since the sealed area had been abandoned inpart because of a history of roof falls, therewas some speculation early in the course ofthe investigation that a roof fall might haveprovided a frictional ignition source. Subse-quent identification of the apparent ignitionpoint and the forces emanating from thatpoint suggested, however, that a roof fallprobably was not the triggering mechanism,leaving lightning or spontaneous combustionas the alternative causes.

Since spontaneous combustion has beendeemed unlikely by most investigators, theyhave focused on attempting to determinehow lightning found its way into the sealedarea. Although various theories have beenadvanced, the facts must drive the theoryrather than the other way around, and somekey facts have yet to emerge. Therefore, untilthe issue of the proximate cause has beendefinitively resolved, speculating is pointless.

That said, however, three questions clearlyrelevant to both the cause and the conse-quences of the explosion must be addressed:

• Prevention: Did the Sago Mine’s owner,International Coal Group (ICG), do every-thing required by law and regulations toprotect the mine against lightning strikes?

• Containment: Was the sealed area properlysealed, and were the seals used strong enoughto contain a substantial explosion?

37

• Propagation: Was propagation of theexplosion minimized by effective rockdustingoutby the sealed area?

From testimony provided to investigators andat the public hearing in Buckhannon on May2-4, 2006, it is clear that ICG failed to properlyground the mine’s electrical power infrastruc-ture in its entirety, and failed to install light-ning arrestors at some key locations as re-quired by federal regulations. Pending furtherinvestigation, the question of whether thesefailures directly contributed to the explosionand subsequent loss of life remains to beresolved, but there is no question that theyrepresent serious failures of mine manage-ment.

As to the seals themselves, in the early stagesof the investigation much attention was paidto the question of whether they were properlyconstructed. That question has not beenentirely resolved and may never be, sincemost of the seals were pulverized. But tosome extent the focus on this question had theeffect of distracting attention from the largerand more relevant question, which is whetherthe Omega blocks that were used to constructthe seals were, in fact, capable of withstand-ing a severe explosion originating inby theseals. Although this question is explored inmore detail in section 5 of this report, it isimportant to note here that this question doesnot go to the adequacy of ICG’s managementof the Sago Mine. The mine operator installed,and relied upon, a type of seal approved forthe purpose by the federal government.

Thus the crucial question with regard to theseals at Sago is whether the criteria used todetermine the acceptability of Omega blocksas explosion-proof bulkheads were appropri-

ate and adequate. The evidence available atthis point suggests that the test criteria wereinsufficiently comprehensive and failed to testthe ability of Omega blocks to withstand andsurvive dynamic pressure from an explosionoriginating inby the seals. This leads to theconclusion that Omega blocks should neverhave been approved as explosion-proof seals,because it appears highly unlikely that, nomatter how they were constructed, they couldhave contained an explosion such as the onethat occurred in the Sago Mine on January 2.

Addressing the last of the three cause-relatedquestions raised above, however — thequestion of whether propagation of the blastwas effectively contained by adequaterockdusting outby the sealed area — it needsto be said that the explosion on January 2might well have resulted in even greater lossof life but for the fact that investigators foundthe mine to have been well rockdusted priorto that date. Some investigators, in fact,consider Sago to have been exceptionally wellrockdusted, apparently reflecting the effortsof mine managers to improve safety condi-tions at the mine in the wake of meetings heldin late 2005 with MSHA representatives.

Rockdusting — the systematic application ofpowdered limestone to mine surfaces —suppresses coal dust, a substance even moreexplosive than methane. Deprived of coaldust as a mechanism to propagate the forcesgenerated by the initial ignition, the explosionof January 2 had lost much (although cer-tainly not all) of its power before it reachedthe One Left crew at their location furtheroutby the sealed area. They experienced theblast mainly as a powerful rush of air anddebris. A coal dust explosion, in contrast,might have engulfed the entire mine, killing

38

everyone or nearly everyone inby the portal.Moreover, deprived of coal dust, the explo-sion did not ignite a mine fire, which if it hadspread could have made escape for the OneLeft crew problematic or impossible.

Accordingly, subject to development ofadditional and definitive evidence, we believe

that the question of what caused the SagoMine disaster has a three-part answer:

• Lightning probably caused the explosion.More remains to be learned, however, abouthow lightning made its way into the sealedarea where the methane ignition took place,and, depending on how that question is

LIGHTNING RELATED EXPLOSIONS IN COAL MINES (1993-2006)

An examination of the 9 known or suspected lightning strikes in 5 U.S. coal mines, since 1993, is an unsettlingexercise. In contrast to other MSHA investigation reports which provide a measure of engineering certainty, theseare more speculative, containing phrases like “probable cause was lightning” or ‘it is reasonable to conclude thatlightning may have ignited methane.” We need prompt research on the role lightning plays in underground mineexplosions and ways to prevent it.

Mine Name/ Operator & Coal Seam Date & Time Description of SealMSHA ID # Location & Depth of Explosion

1 Mary Lee Drummond Mary Lee 8/22/93 Cementitious foamNo. 1 Company, Inc. Seam 4:30 P.M.01-00515 Goodsprings, AL 600/700

2 Oak Grove U.S. Steel Mining Blue Creek 4/5/94 5 solid block concrete,01-00851 Company, Inc. Coal Seam; 10:50 P.M. mortared; 33

Adger, AL 1100’ cementitious foam seals

3 Gary No. 50 U.S. Steel Mining Pocahontas 6/9-6/16/95 TEK foam, 4 ft. thick seals46-01816 Company, Inc., No. 3;

Pineville, WV 600/700’

4 Oak Grove U.S. Steel Mining Blue Creek 1/29/96 5 seals, 6 ft. thick; 3801-00851 Company, Inc. Coal Seam; cementitious foam by

Adger, AL 1100’ Celtite Div./Fosroc, Inc.

5 Mine No. 1 Oasis Contracting, Cedar Grove 5/15/96 Micon 550, 6” solid46-07273 Inc. Quinlan, WV Seam;390’ 4:00 A.M. concrete blocks as framing

walls for 16” polymer groutand aggregate inner core

6 Mine No. 1 Oasis Contracting, Cedar Grove 6/22/96 Micon 550, 6” solid46-07273 Inc. Quinlan, WV Seam;375’ 8:00 P.M. concrete blocks as framing

walls for 16” polymer groutand aggregate inner core

7 Oak Grove U.S. Steel Mining Blue Creek 7/9/97 Cementitious foam seals01-00851 Company, Inc. Coal Seam; 8:15 P.M. by Celtite Div./Fosroc Inc.

Adger, AL 1100’

8 Gary No. 50 U.S. Steel Mining Pocahontas 5/18/01 Micon 55046-01816 Company, Inc. No. 3; 5:09 P.M.

Pineville, WV 600/700’

9 Sago International Coal Middle 1/2/06 40” thick Omega block46-08791 Group, Inc., Kittanning 6:26 A.M. seal

Tallmansville, WV Seam;300’

➔

➔

➔

➔

➔

➔

➔

➔

➔

➔

39

(chart continues...)

answered, specific steps must then be takento protect other underground mines andminers against the risk of lightning-relatedmethane ignitions.

• The seals that should have helped containthe explosion were not sufficiently explo-sion-proof and never should have beenapproved for that purpose. The catastrophicfailure of the Omega block seals clearly madethe explosion far more lethal than wouldotherwise have been the case.

• Effective rockdusting of the mine prior toJanuary 2, 2006, helped contain the explo-sion and saved lives. It is important tounderscore this point, both to give creditwhere credit is due and to highlight theimportance, on an industrywide basis, ofguarding against mine emergencies by takingthe kinds of basic preventive measuresrequired both by law and by experience-based risk assessment and best practices.

Effect on Seals Magnitude Suspected Cause Effect onof Strike Miners

2 seals blown out (-)21.7 ka Explosion was a result of a lighting No personneland fan shaft cap (electrification) strike at the vent pipe injureddamaged pipe on the cap of 70 North fan shaft

2 seals blown out (+)84.8 ka Lightning strike close to abandoned No personneland 1 damaged cased bore hole ignited methane injured

No seals blown out (-)90.2 ka; Lighting strike could have caused No persons7 strikes explosion; no direct path identified in mine

5 seals destroyed (+)55.8 ka Lightning strike near any wells could No personnelwith pieces blown couple energy onto any of the metallic injured80’ to 100’ casings, travel down to an entry

in the sealed areas

3 seals blown out, (+)59 ka Probable cause was either lightning No persons1 partially blown strike or friction/ignition from a in mineout. (Non-producing roof fall mine;seals leaked)

Not reported (-)50 ka Probable cause was either lightning No personsstrike or friction/ignition from a in mineroof fall

3 seals destroyed (-)142.2 ka Path into the sealed areas undefined No personneland 1 severely injureddamaged

Seals intact Unknown Lightning strike confirmed at time Force confinedand vicinity of explosion; reasonable by seals;to conclude that methane ignition no loss of liferesulted when an explosive mixturewas ignited by lightening

All 10 seals pul- (-)101ka Lightning strike may have caused 1 killed initially;verized; no seal explosion; no path identified 16 escaped; 12material attached or confirmed trapped,eventuallyto roof, rib or floor only one survivor.

Source: MSHA investigation reports, with the exception of the Sago entry.

➔

➔

➔

➔

➔

➔

➔

➔

➔

➔

40

SOURCES

1 Monte Hieb, memorandum, 7/5/06.

2 This event was originally thought to have occurred at 6:31 a.m. However, state investigators deter-mined that the computer clock at the mine was 4 minutes 56 seconds fast. Adjusting the clock toUniversal Time (atomic clock time), they found that the CO monitor alarmed at 6:26:35, precisely thesame time that the lightning strikes were recorded by Global Positioning System (GPS) clocks, whichare also set to Universal Time.

3 This reading appears to have actually exceeded the monitor’s maximum level (50ppm). When it pegsat that level the monitor is designed to shut itself down and must be reset.

4 Monte Hieb, memorandum, 7/5/06.

5 Monte Hieb, memorandum, 7/5/06.

41

5 Why did the seals fail?

‘The current seal construction regulations… assume that an explosionoccurring in [a mined-out and abandoned area] will not be stronger than20-psi pressure. However, if a large flammable gas volume exists in the[sealed area], the resulting explosion pressure can be >20 psi.’

— National Institute for OccupationalSafety and Health, 20011

The explosion in the Sago Mine originated ina recently abandoned and sealed area. Intheory, the seals constructed to isolate theabandoned area from the adjacent workingsections of the mine should have withstoodthe blast — unless the ignition of methaneinby the seals generated forces so powerfulthat no barrier, regardless of the materialsused and the method of construction, couldhave contained the explosion.

That does not appear to have been the case.Instead, what appears to have happened isthat the supposedly “explosion-proof” sealsfailed because they were constructed of amaterial — Omega blocks — that shouldnever have been approved for the purpose ofcontaining an explosion originating in asealed area.

On May 12, 2006, Acting Director James M.Dean of the West Virginia Office of Miners’Health, Safety and Training imposed a state-wide moratorium on the use of Omega blocksas explosion-proof seals in underground

mines. Eight days later, on May 20, 2006, anexplosion in the Kentucky Darby No. 1 Minein Harlan County, Kentucky, blew out a sealconstructed of Omega blocks, killing fiveminers in the explosion and its aftermath. OnMay 22, 2006, the federal Mine Safety andHealth Administration imposed a temporarynationwide moratorium on the constructionof alternative seals such as Omega blocks,requiring mine operators wishing to sealworked-out areas to revert to the practice ofusing solid concrete-block bulkheads, pend-ing a reassessment of the structural integrityof alternative seals.2

West Virginia’s decision to bar the use ofOmega blocks as seals was appropriate andshould be made permanent. Moreover, thestate should join with the United MineWorkers of America, which has opposed theuse of Omega blocks as seals for many years,and with other concerned parties to spur theappropriate agencies to conduct a compre-hensive review of the entire history of thedevelopment of explosion-proof criteria for

42

seals used in the mines of the United States.These criteria appear to be severely inad-equate and in urgent need of revision. In themeantime, Omega blocks should not beconsidered explosion-proof regardless ofhow they are constructed, and the currenttemporary nationwide moratorium on theiruse in underground mines for purposes otherthan as ventilation stoppings should bemade permanent.

Although a comprehensive assessment of thedevelopment of the criteria currently govern-ing the construction of explosion-proofbulkheads in U.S. mines is beyond the scopeof this report, a brief discussion is warrantedin order to explain the need for furtherinvestigation and action.3

The worst mine disaster in the nation’shistory occurred in 1907, when an explosionin a coal mine at Monongah, West Virginia,claimed 362 victims. In the wake of thatcatastrophe and several others that occurredover the next two years, Congress establishedthe federal Bureau of Mines in 1910. Althoughthe Bureau was initially given no enforcementauthority (strong federal mine legislation wasnot enacted until 1969), the Bureau did havea mandate to investigate the causes of fatalaccidents and to make recommendations forimprovements in safety and health protection(which mine operators were free to adoptor ignore).

The Bureau’s research on sealing abandonedmine areas dates back to 1914, when theagency began conducting explosion tests in itsnew experimental mine. Over the years,sometimes in cooperation with similar agen-cies in other mining countries such as Ger-many and the United Kingdom, the Bureau

undertook numerous studies of explosionsand developed strategies to prevent andcontain them.

Early on, experts were able to arrive at aconsensus view regarding the basic standardsneeded to contain explosions. In 1921, forexample, the U. S. government, in adoptingcriteria for explosion-proof bulkheads be-tween coal mines on government-ownedlands (the only sphere where the governmenthad the power to regulate at the time), re-quired that such stoppings be able to with-stand an explosion generating a pressure of 50pounds per square inch (psi).

Similarly, commissions investigating disastersin the United Kingdom in the 1930s and in1960 concluded that mine explosions arelikely to develop pressures ranging from 20 to50 psi, and determined that a 50-psi standardwould provide a reasonable margin of safety.Later studies of explosions in Poland andGermany led researchers there to concludethat bulkheads should be designed to with-stand higher pressures, and that a 72-psistandard would be advisable.

Despite these findings, however — anddespite subsequent adoption of 50 psi and insome cases more rigorous standards inmining countries such as Canada, Australia,and South Africa — the Bureau of Minesadopted a much lower standard for theUnited States.

The Federal Coal Mine Health and Safety Actof 1969 — enacted in the wake of the explo-sion that killed 78 miners at a ConsolidationCoal Company mine near Fairmont, WestVirginia, in November 1968 —requiredbulkheads to be “explosion-proof.” The

43

Bureau of Mines was then charged withdeveloping appropriate standards to give thatterm meaning. Notwithstanding the increas-ing adoption of the 50-psi (or higher) stan-dard elsewhere, the Bureau published a singlereport in 1971 that made the case for a 20-psirequirement — a standard subsequentlyadopted in federal regulations [30 CFR75.335] that remain in effect today.

That 1971 report (Explosion-Proof Bulkheads:Present Practices, U.S. Bureau of Mines Reportof Investigations 7581) makes for deeplydisturbing reading in the wake of the SagoMine disaster. The author of the report, thelate Donald W. Mitchell, then Acting Supervi-sor of the Bureau’s Pittsburgh Mining andSafety Research Center, briefly reviewed U.S.and foreign experience with mine explosionsand noted that in test explosions in theBureau’s experimental mine, pressures hadregistered as high as 127 psi. He then stated,however, that pressures 200 feet or more fromthe origin of an explosion seldom exceed 20psi. That statement, while not documentedanywhere in the report, appears to be therationale for his conclusion that bulkheadsmay be considered “explosion-proof” if theycan withstand a static load of 20 psi.4

Mitchell’s report makes no clear distinctionbetween static and dynamic pressures, eventhough the differences can be important.Picture an explosion taking place at oneend of a relatively long, narrow room. Theexplosion, propagating outward into theconfined space, exerts static pressure on thewalls parallel to the direction of travel, anddynamic pressure on the opposite wall. In thisexample, the opposite wall takes the greaterhit. But from the evidence available it appearsthat Mitchell, in arguing for a 20-psi standard,gave little consideration to dynamic pressures.

The Bureau in its experiments appears tohave tested materials for static pressureresistance only — that is, installing them asventilation stoppings would be installed,typically in crosscuts parallel to the likelysource of a blast, rather than as barriers builtin the entry directly across its path. Conse-quently, the tests conducted over the years bythe Bureau of Mines and its successor agen-cies — the Mine Health and Safety Adminis-tration and the National Institute for Occupa-tional Safety and Health — to determinewhether a given material and structure is“explosion- proof” as required by federal lawdo not seem to address the kind of explosionexperienced at Sago.

It should be noted, and emphasized, that wehave not yet had an opportunity to review theapparent limitations in the methodology ofMitchell’s report with experts at MSHA orNIOSH, and we need to be open to thepossibility of having missed something.However, unless that is the case, it is difficultto escape the conclusion that the 20-psistandard — adopted entirely as a result ofthat one 1971 report — provides none of themargin of safety that government explosionexperts were sensibly advocating as far backas 1921, when they proposed 50 psi as astandard for seals survivability.

Why is this important now? It is importantbecause Mitchell’s 1971 report became, for theBureau of Mines and then for MSHA andNIOSH, the standard reference which for thepast 35 years has apparently guided theparameters for the testing and certification ofallegedly “explosion-proof” seal designs andmaterials — including Omega blocks.

The Omega 384 block, manufactured byBurrell Mining Products, is a lightweight

44

fiber-reinforced block consisting of a mixtureof cement, foam, and fly ash. Because astandard Omega block (16 inches by 24 inchesby 8 inches) is three times the size of a singlesolid concrete block but weighs far less, theproduct has become popular in the miningindustry, where it is widely used to constructventilation stoppings. Omega blocks can beeasily stored and transported into the mine onpallets, and stoppings made of the blocks canbe built more easily and with considerablyless risk of musculoskeletal injuries thanwhen working with solid concrete blocks.Omega blocks can also be cut with a handsaw and are incombustible as well as mois-ture-resistant.

In 1990, MSHA was asked to approve Omegablocks as explosion-proof seals. Explosiontests were conducted in October 1990 at theLake Lynn Experimental Mine. Burrell em-ployees constructed four seals, each of whichsurvived a pressure pulse of approximately20 psi. However, as with previous “explosion-proof” tests, the testing was done with theseals constructed in crosscuts, as though theywere ventilation stoppings, rather than builtdirectly in front of and across the source of theblast. Thus the seals demonstrated only theirability to survive a powerful but glancingblow, rather than a full-force, head-on blow.In short, they were not subjected to the kindof test to which Omega blocks would besubjected at Sago nearly 16 years later. Byway of analogy, think of an automobile beingtermed “collision-proof” on the basis ofsurviving a glancing blow at an intersectionrather than a head-on crash at highwayspeeds.

Few in the mining industry — or, for thatmatter, in the ranks of mine safety enforce-

ment — were probably aware of the tests’limitations. To most of us, the knowledge thatthe government will not sign off on some-thing as “explosion-proof” unless it with-stands an explosive force of 20 pounds persquare inch has been good enough. Whatmore would we need to know? And, prior tothis year’s disasters at Sago and Darby, whatwe knew, essentially, was that MSHA hadapproved Omega blocks in 1992 as alternativeseals, and that they were considered explo-sion-proof as long as the construction of theseals was done in accordance with the waythey had been tested as well as in accordancewith the specific, MSHA-approved ventila-tion plan for the mine installing them.

But in reality that was not good enough,because, as we now know, the tests them-selves were not good enough. They did nottest the ability of seals built of Omega blocksto survive a direct hit — a force of 20-psi orhigher — from inby the seals.

In hindsight, it is difficult to understand whysuch tests were never conducted — never,that is, until mid-June 2006, six months afterthe Sago disaster. (In the latest tests, theOmega-block seals failed.) If the assumptionwas that explosions would not originate inabandoned and sealed areas because therewould be no ignition sources in such areas,that is contradicted by long experience. Forexample, a roof fall in a sealed area couldtrigger a frictional methane ignition if itoccurs while a methane buildup in the sealedarea is in the explosive range (5 to 15 percentof the mine atmosphere). Moreover, evidencehas been accumulating for several years thatlightning can ignite methane in an abandonedand sealed area, triggering an explosion.

45

Between 1993 and 1997, seven explosions inabandoned areas of underground mines inAlabama and West Virginia were found tohave been caused by lightning. MSHA inves-tigators found that, in most cases, sealsconstructed of various materials had beenpartially blown out by the blasts. Fortunately,in all of these cases the lightning strikesoccurred in remote areas of the mines; theexplosions were not severe; and no minerswere working directly in their path or inaffected areas.

In the wake of these incidents, NIOSH issueda bulletin in 2001 to alert the mining industryto the apparent hazard of lightning-triggeredignitions, noting that in at least one case theforce of the explosion clearly exceeded 20 psi.The bulletin further noted:

“The current seal construction regulations,which relate to all suitable underground sealdesigns, assume that an explosion occurringin the gob [mined-out and abandoned area]will not be stronger than 20-psi pressure.However, if a large flammable gas volumeexists in the gob, the resulting explosionpressure can be >20 psi.”5

Unfortunately, NIOSH concluded that iflightning as a potential source of ignitioncould not be eliminated, the only alternativewould be to eliminate flammable concentra-tions or reduce the volume of the flammablemixture present in the sealed area (by takingsteps such as minimizing pressure differen-tials across seals to reduce air leakage). Othersteps, such as eliminating wires or otherconductors and removing potential ignitionsources such as old batteries from sealedareas, were also recommended. But neitherNIOSH nor MSHA, despite their acknowl-

edged awareness that sealed-area explosionscould generate more than 20 psi, recom-mended reconsidering the 20-psi standard for“explosion-proof” bulkheads in the light ofaccumulating experience.

The decision by MSHA and WVOMHST inMay 2006 to impose a moratorium on the useof Omega blocks as seals comes too late forthe victims of the Sago Mine disaster. But itwould seem to indicate a recognition, even ifnot explicitly acknowledged, that both the 20-psi standard and the explosion-proof testingregime are long overdue for reassessment.State and federal authorities must now worktogether — and must be willing to draw uponthe experience of other mining countries andthe more stringent standards they have longsince adopted — to ensure that this urgentpriority is addressed without further delay.

In the wake of the Sago Mine disaster, muchattention was paid to the question of whetherthe Omega block seals at Sago were properlyconstructed. That would seem to be a largelyirrelevant question, even something of adistraction, since the evidence stronglysuggests that they would have been de-stroyed regardless of how they were con-structed. The more basic problem is with astandard for “explosion-proof” that is clearlyinadequate, both in terms of the standarditself — 20 psi — and the tests traditionallyused to determine whether a given materialcan meet that standard.

Standards for explosion-proof bulkheads inWest Virginia and throughout the UnitedStates must now be brought into line withglobal standards, which suggests setting thebar at pressures significantly higher than 20psi, and remedial steps must be taken to

46

ensure that all seals now in use are strength-ened, if necessary, within a specific andenforceable timeframe.6 Where appropriate,on a case-by-case basis, boreholes should bedrilled into abandoned areas to ventilatethem. Other steps such as inerting sealed

areas by injecting nitrogen may also bewarranted and should be explored with alldeliberate speed.

For now, only seals constructed of solidconcrete blocks (also known as Mitchell-

An explosion-proof seal

Staggered block arrayto strengthen pilaster

Pilaster center

Transverse pattern of laying block

The diagram above shows a solid concrete block seal (also known as a Mitchell-Barrett seal). Forextra strength the blocks are staggered — laid both lengthwise and crosswise — and all jointsare mortared. To provide strength at the edges of the seal, the seal is hitched into the coal rib.When a seal is to be constructed above a specified height, it must be further strengthened byincluding a pilaster — an extra row of staggered blocks in the center as shown. Construction ofsuch a seal is time-consuming, labor-intensive, requires masonry skills, and can involve riskssuch as musculoskeletal injuries while handling the heavy blocks and the possibility of beinginjured by a rib roll while hitching the rib. Consequently, alternative seals were developed andhave become popular. Some, such as cementitious materials poured or pumped into framesfilling the entire void between floor and roof, may have bulk and strength sufficient to withstandsevere explosions. But Mitchell-Barretts should be thought of as the gold standard.

47

Barrett seals) or seals of demonstrated equiva-lent strength should be used for constructionof seals in West Virginia and throughout thecountry. Mitchell-Barrett seals (see diagram)have successfully withstood explosionsgenerating pressures far greater than 20 psiand offer a much higher level of protectionthan afforded by Omega blocks and otheralternative seals.

In a related matter, research needs to beconducted to better understand the impactthat variations in mining height and abruptchanges in the cross-sectional area of mineentries in a sealed area may have on thevelocity of an explosion. At Sago, for example,bottom mining had been done in the areawhere the explosion occurred. We need toknow the impact that bottom mining mayhave had on the strength, velocity, anddirection of the explosive forces resultingfrom the methane ignition. As is well known,the velocity of water increases when it movesfrom a large pipe to a smaller pipe (an actionknown as the Venturi Effect). Because air actsin the same way as water, an explosive forcewhich is initiated in a large space is thenforced into a smaller space can be expected toresult in a similar increase in velocity. Inaddition, turbulence that is created when ablast wave passes across entry obstructions orsudden mining-height changes increases

mixing of stratified mine gases and canenhance propagation effects.7

Computer modeling systems are availablethrough institutions such as Sandia NationalLabs, Naval Surface Warfare Center, ArmyCorps of Engineers facilities, and others. Suchtests should be funded and undertakenimmediately in order to better understandthe explosive forces which could be poten-tially created in coal mines by large andnumerous changes in entry height due tobottom mining.

In addition to maintaining the current mora-torium on the use of Omega blocks as seals, acomprehensive survey of the numbers andlocations of Omega-block type seals currentlyin use should be completed, and plans toretrofit or otherwise accomplish the construc-tion of explosion-proof seals within a specifictimetable should be pursued, with each WestVirginia mine operator responsible for devel-oping plans subject to approval by the WestVirginia Office of Miners’ Health, Safety andTraining.

Until measures such as these have been taken,every sealed area in every underground coalmine in West Virginia and throughout theUnited States should be considered a poten-tial time bomb — and treated accordingly.

SOURCES

1 NIOSH Technology News 489, Reducing the Danger of Explosions in Sealed Areas (Gobs) in Mines, May2001.

2 For more information, see: www.msha.gov/regs/complian/PIB/2006/pib06-14.asp

3 A review of the statutory and regulatory history of the present standard is available at www.wju.edu/sago/def.asp. See Memorandum for the File, MSHA 20 psi Seal Standard.

48

4Mitchell, Donald W., Explosion-Proof Bulkheads: Present Practices, Report of Investigations 7581, U.S.Dept. of the Interior, Bureau of Mines, 1971.

5 NIOSH Technology News 489, Reducing the Danger of Explosions in Sealed Areas (Gobs) in Mines, May2001.

6 Monte Hieb, memorandum, 7/11/06.

7 Monte Hieb, memorandum, 7/5/06.

49

6 Did the miners’SCSRs fail?

‘… The mine filled quickly with fumes and thick smoke… The firstthing we did was activate our rescuers, as we had been trained. At leastfour of the rescuers did not function. I shared my rescuer with JerryGroves, while Junior Toler, Jesse Jones and Tom Anderson sought helpfrom others. There were not enough rescuers to go around.’

— Randal McCloy, Jr.1

When Randal McCloy, Jr., the only survivoramong the miners trapped in the Two Leftsection of the Sago Mine on January 2, recov-ered to the point of being able to release astatement to the families of his co-workers onApril 26, he delivered terrible news. The crewhad survived the explosion, only to discoverthat their first line of defense — the self-contained self-rescuers (SCSRs) they carriedinto the mine with them every single workingday — had failed them. They could not makeall of their SCSRs work. Without enough fullyfunctioning SCSRs to go around, there was noway, as they understood the situation, for allof them to make it safely out of the sectionand back to where there might possibly berespirable air.

When Mr. McCloy in his continuing recoverywas able to be interviewed by state andfederal investigators on June 19, he amplifiedon his earlier statement, explaining how hestruggled with the non-functioning SCSRsand could not make them work. He describedsharing his own SCSR with his roof-bolting

buddy, Jerry Groves, and how he struggledwith Mr. Groves’s SCSR: “I fought with it forI don’t know how long, trying to mess withthat valve, blow air through it or anythingI could do, but nothing would work.”2

The apparent failure of some of their SCSRs togenerate oxygen may have doomed theminers trapped on Two Left. Without enoughfunctioning SCSRs to go around, they couldnot have tried to walk out of the mine, know-ing that they would making their waythrough lethal concentrations of carbonmonoxide and that any miner without aworking SCSR would rapidly succumb.And it is impossible to imagine any of them,in that situation, simply walking away fromtheir fellow miners.

Instead, according to Mr. McCloy, they triedto make their way together to where they hadleft their mantrip — yelling to keep track ofeach other in the blinding smoke — but hadto abort that escape attempt when theyencountered thick dust and swirling smoke

50

and what they believed was impassabledebris. “We couldn’t escape the smoke,” Mr.McCloy recalled. “There was nowhere to gobecause it just lingered everywhere, youknow, just everywhere you went.”3

So the miners returned together to the sectionface, where they hung a curtain as a barri-cade, began trying to signal their location tothe surface by beating on roof bolts with asledgehammer, and shared the limited num-ber of SCSRs that did seem to be working.

Meanwhile the miners on the One Left crew,further outby the explosion, were makingtheir way into the escapeway and out of themine after being engulfed in a rush of air,dust, and debris. Seven of them donned theirSCSRs. From their subsequent testimony toinvestigators it appears that, with one pos-sible exception, all of their SCSRs worked,although at least two miners had somedifficulty using them.

Mine Safety and Health Administration(MSHA) investigators subsequently recov-ered all of the SCSRs from the Sago Mine,including the Two Left section, had themtested by the National Institute for Occupa-tional Safety and Health (NIOSH), andreported that all were in working order. Thisseems to raise the question: Who is right?The miners who believed that their SCSRsfailed, or the laboratory technicians whoexamined those same units and found themto be operational?

That is the wrong question. The right ques-tion is not whether SCSRs work well underideal conditions in a NIOSH laboratory.The question is whether they work in themine, in an emergency, when they’re desper-ately needed.

To address that question, we must look first athow an SCSR is typically used, and then atthe quality of the training provided to theminers who must rely on them.

The situation on Two Left in the momentsimmediately after the explosion tells us all weneed to know about the kinds of conditions inwhich an SCSR is likely to be used. The minesuddenly fills with noise, smoke, and fumes.It’s pitch-black. The smoke and dust are denseand swirling. A miner’s cap lamp — muchmore powerful than an ordinary flashlight —cannot penetrate the murk. The mine atmo-sphere becomes terrifyingly foul and staysthat way. Breathing immediately becomesdifficult. Miners quickly lose contact witheach other and can easily become disoriented.They know, instantly, that they are in adesperate life-and-death situation in whichseconds count, and that the SCSRs they carryare their first and possibly last line of defense.So their first priority is to activate their SCSRs— under just about the worst kind of stressimaginable.

The Sago miners were equipped with the SR-100, an MSHA-approved SCSR manufacturedby CSE Corporation in Monroeville, Pennsyl-vania. Like other SCSRs approved by MSHA,the SR-100 is rated to provide a minimum ofone hour of breathable air while the userbreathes under exertion, such as while escap-ing a mine, and in laboratory tests has pro-vided approximately four hours of air at rest.The belt-wearable SR-100, which weighsabout six pounds and costs about $680 perunit, is currently the most widely used SCSRin U.S. mines.

Problems have been experienced with the SR-100 over the years. In 1998, for example,

51

MSHA issued an alert to the mining commu-nity after finding that some SR-100s emittedhigher than normal levels of carbon dioxidein the air supplied to the user. Although theSCSRs still functioned, MSHA warned thatminers using them might find themselvesbreathing uncomfortably fast and moredeeply than expected, possibly causing themto become frustrated with the SCSR anddiscarding it. CSE suggested that the devicesmost likely to produce higher than normalcarbon dioxide levels were those that hadbeen exposed to excessive shock and vibra-tion over an extended time.

Similarly, MSHA issued an alert in 1999 afterfinding that a small number of SR-100smanufactured from 1991 to 1993 had deterio-rated hoses. CSE pledged to replace any suchunits, and subsequently modified the SR-100’s hose design. MSHA urged mine opera-tors to perform regular, thorough inspectionsof their SCSRs, remove from service any unitswith detectable damage, and replace olderunits. MSHA also advised placing extraSCSRs on each mining section.

Given the widespread use of the SR-100,however, problems have been relatively rare,and it has generally been considered a reliableunit. But general reliability is no comfort to aminer suddenly facing a life-threateningemergency. At that instant, the miner’s onlyquestion is: “Can I get this thing on in a hurryand will it work?”

The answer to that vital question is not assimple as we would like. A number of vari-ables are involved, the most important ofwhich is the quality and frequency of thetraining that the miner may have receivedprior to needing the SCSR in an actualemergency.

Activating and donning the SR-100 requirespractice and clear thinking. The miner re-moves the SCSR from its pouch. Holding it inhis hand, he kneels on the ground and re-moves his hard hat so that he can shine hiscap lamp on the area immediately in front ofhim. He opens the SCSR by lifting the latchon top, which supposedly allows him toremove both the top and bottom protectivecovers. In practice, removing the covers canbe difficult, as the manufacturer acknowl-edges in its instructional materials. The minerthen removes the safety goggles packed in theunit, looping them over his wrist so he won’tlose them in the murk and mud. He takes theneck strap out and puts it over his head,taking care not to put the unit on backwards,which would make it difficult to use themouthpiece and nose clips properly. He pullsan orange actuator tab downward to activateoxygen (the unit contains a small oxygen

Donning an SCSR is awkward and uncomfortableunder the best of circumstances — let alone in asmoke-filled mine.

52

reserve to help inflate its breathing bag). Heextends the breathing hose, removes a plugfrom the mouthpiece, and inserts the mouth-piece in his mouth, biting down hard on it tohold it in place. He pulls the two nose padsapart and affixes them to his nose, completelyclosing both nostrils. Although it iscounterintuitive, he then exhales into the unitto get its rebreathing mechanism going. Heputs on the safety goggles. He adjusts theneck strap so that the SCSR rests on his chest.He then fastens the waist strap tightly to keepthe SCSR from swinging loose as he moves.He puts his hard hat back on his head, standsup, and starts planning his escape. (He can’tvoice-communicate with other miners whileusing his SCSR, so he must try to maintainphysical or visual contact.) If everything goesperfectly, he gets his SCSR up and running inabout 30 seconds.

It isn’t easy, even if you’ve been thoroughlytrained and are young and in great shape. Ifyou’re older, and not in such great condition,it’s harder. If your training has been superfi-cial, and if you haven’t been retrained re-cently, it’s harder still. If you wear eyeglassesand they’ve been blown away and you’reblinded by smoke and grit, and you’re chok-ing, gasping, and at the brink of panic, it’sharder yet. Several of the One Left minersdiscovered that they just about could not do itwithout help from other miners.

Moreover, the challenge doesn’t end withdonning the SR-100. It is possible to“overbreathe” the unit — in which case it willfeel to the miner as though he is trying to suckair through a crimped drinking-straw insteadof breathing through a hose. As ICG’s assis-tant safety director, Johnny Stemple, ex-

plained while demonstrating an SR-100 at theMay 2-4 public hearings in Buckhannon:

“The [SR-100] will provide oxygen at a certainrate. And you can overbreathe those rescuers.You can breathe faster than they can provideoxygen. If you’re in a mode where you’removing very fast, very quickly, you canoverbreathe those units. You have to slowyour breathing rate down or slow your pacedown to allow those units to produce oxygenat — or to allow your body to take in theoxygen as fast as [the] unit is producing theoxygen. It’s not a unit that provides pressure,so when you activate the unit, there’s notoxygen coming out of it. You have to breatheinto it and breathe out of it, or suck out of it,to get oxygen. And you can outbreathethose units.”4

Stemple’s description, while accurate, doesnot mention another complication that canarise if a miner has not been thoroughlytrained on the SR-100. In fact, the unit doesgenerate oxygen very briefly when it is firstactivated — “a few puffs,” as some trainerssay — so it is possible to get the impressionthat it is working comfortably at first but then,after a couple of breaths, breathing may seemto become much more laborious. Someonethoroughly familiar with the unit wouldunderstand that it is still functioning, justdifferently — because, after those first fewpuffs, the unit needs to have the user activelyexhaling into the rebreathing mechanism inorder to be able to convert exhaled carbondioxide into oxygen. However, someone whohas never had to use the unit could easily getthe impression that it has suddenly stoppedworking, or that breathing through it hassuddenly become unnervingly difficult.Instead of making escape from a smoke-filled

53

mine easier, the SCSR is making it evenharder — or so it may seem. At that pointeven a veteran miner might be tempted tomake the potentially fatal decision to give upon using the SCSR, hoping somehow to getalong without it.

We do not know whether this is exactly whathappened to the miners on Two Left. But wedo know what matters — that some of themwere convinced that their SCSRs were notworking. We do not for a moment believe thatthe miners trapped on Two Left were incom-petent. We are sure, however, that they hadnot been adequately trained in the use of theirSCSRs — and there is also the possibility thatsome of the SCSRs they relied on had beenbadly battered from long exposure to therigors of daily transportation into and out ofthe mine.

From testimony provided to MSHA and stateinvestigators and at the May 2-4 publichearing, we can be reasonably certain thatthose Sago miners who received SCSR train-ing after International Coal Group acquiredthe mine were, in most instances, requiredonly to demonstrate to an instructor that theyknew how to don the unit, while in othercases they may have simply been shown howto do it. (And in at least one case a Sago minertestified that he was asked to sign a slipcertifying that he had been retrained on theSCSR when in fact he had been absent thatday.)5 It should be said, however, that thetraining provided by ICG was probablytypical for the mining industry at this point intime.

The Sago miners did not actually breathe intoa unit, because that is not part of the requiredtraining. And there can be no doubt that they

did not experience in training the kinds ofchallenges, including overbreathing, that theywould face in a real crisis.

The reason why there can be no doubt aboutthat is that ICG has not purchased the SR-T, a$220 training unit marketed by CSE designedto “give the user a chance to experience thesensations that are common with the SR-100[including] breathing resistance and tempera-ture sensations.”6 As CSE notes in its market-ing literature, a miner wearing the SR-T in aworkplace setting can get a sense of what it islike to use the SR-100 under stress: “Higherwork rates will increase breathing resistanceand inspired temperature.”7 That would bevaluable experience to store in memory.

Again, it should be noted that ICG is by nomeans alone in opting not to purchase thisCSE training model. As far as we have beenable to determine, CSE has sold few if any ofthese units to U.S. coal producers, whounderstandably prefer the less expensiveapproach of using old SR-100s as trainingmodels. (The SR-100 has a storage life of 10years, after which units must be discarded orused for training.) The manufacturer reports,however, that the SR-T is in widespread usein Australia, not by government regulationbut by a “duty of care” agreement within themining industry that views thoroughlyrealistic SCSR training as a basic safetyimperative.

We need that attitude in West Virginia andthroughout the coal industry in the UnitedStates. Whether it would have saved the liveslost on Two Left is something we cannotknow for sure. What we can say is that morethorough and realistic SCSR training — andat more frequent intervals than the annual

54

retraining required by the regulations thatwere in effect on January 2, 2006 — wouldimprove the odds that a miner can count onhis or her SCSR in a crisis. And improvingthe odds is a large part of what survival isall about.

The fact that four of the SCSRs relied on bythe Two Left miners did not work for them —leaving them with not enough SCSRs to goaround — is tragic. The fact that NIOSH’ssubsequent tests found all of the SCSRs fromTwo Left to be in working order compoundsthe tragedy, because if those tests were accu-rate they can only mean one thing: that theTwo Left miners, even though most of themwere veterans, had not been sufficientlyfamiliarized with the challenges of using anSCSR in an emergency setting.

Going forward, it will be necessary, at mini-mum, to do on an industrywide basis what afew companies already do: train miners inrealistic settings, including special rooms ortrailers that can be filled with (harmless)Hollywood smoke. Every miner needs andshould have that level of SCSR training,repeated frequently enough to retain it inmemory. As Joe Ryan, a veteran member ofthe One Left crew, described the experience ofstruggling with his SCSR on January 2:“We’ve heard it so many times, we could do itin our sleep. But then, when you have to do it,it’s a different — it’s really, really....”8 Hedidn’t finish the sentence; he didn’t have to.

In addition to immediately improving train-ing, we need to address the urgent question ofwhether the SCSRs currently in service can becounted on to work in practice. The WestVirginia Office of Miners’ Health, Safety andTraining is attempting to address that ques-

tion by requiring mine operators to inventorySCSRs and report problems on a quarterlybasis, effective August 15, 2006.9 This shouldbe a useful development. However, SCSRsshould also be field-tested. The best way to dothat would be to give federal and state inspec-tors the authority to ask individual miners onworking sections — at random, withoutforewarning, and on a voluntary basis — todon their SCSRs and walk out of the mine, orat least far enough to get a good feel forbreathing through the unit. Since the unit willthen have to be replaced, and because thissafety exercise will also be seen as downtimeand lost productivity, many coal operatorscan be expected to vigorously oppose thisproposal. But some will surely recognize thatas we absorb the lessons of the Sago Minedisaster — in which the initial explosionseems to have been followed by one tragedyafter another — this is the very least we cando to prevent a recurrence. For those optingto invest in and use training-model SCSRs,the random testing requirement could bewaived.

Another necessary step is to reduce the 10-year life expectancy currently permitted forSCSRs. That time period, we believe, isarbitrary and not based on any laboratorydeterminations. To create a greater margin ofsafety — including guarding against the riskof malfunction from the repeated wear andtear that SCSRs experience as they are trans-ported in and out of mines daily — SCSRsshould be withdrawn from use after a shorterperiod of service.

Moving beyond these interim measures, wemust immediately take steps to acceleratedevelopment of the next-generation SCSR.The SR-100 and similar units produced by

55

other manufacturers represent technologiesthat are more than two decades old. Despitetheir obvious performance limitations andincreasing obsolescence, there has been noserious effort to fast-track the research,development, and deployment of the nextgeneration. MSHA and NIOSH have not gonemuch beyond workshop conceptual discus-sions, and no one has stepped forward toassist researchers, inventors and manufactur-ers interested in getting a 21st-century prod-uct into the hands of 21st-century miners.

This must change. Several next-generationconcepts are worthy of support, including ahybrid SCSR which in less toxic environmentscan convert to functioning as an air-purifyingrespirator, and a dockable SCSR that wouldpermit supplementary units to be docked tothe main unit while in use, eliminating theneed for the wearer to switch units. In addi-tion, a prototype two-hour SCSR unit devel-oped by OX-GEN Flexible Life Systems, Inc.,and displayed at the International MiningHealth and Safety Symposium in Wheeling,West Virginia, on April 20-21, 2006, showspromise and merits further support andtesting. The unit can be stored in portable 48-unit caches kept within 50 feet of miners onworking sections, and has a comfortable half-face mask, rather than a mouthpiece, with a

SOURCES

1 Randal McCloy, Jr., “To the Families and Loved Ones of My Co-Workers, Victims of the Sago MineDisaster,” April 26, 2006, p. 1.

2 Randal McCloy, Jr., testimony, June 19, 2006, p. 33.

3 Randal McCloy, Jr., testimony, June 19, 2006, p. 47.

voice transmitter that allows communicationwhile breathing. This represents an obviousadvance that is worthy of further support.

While there is a clear need to push ahead withdevelopment and testing of next-generationSCSRs, there has been a vacuum of leadershipat the national level. This, too, must change,but West Virginia does not need to wait forthat to happen. The state has ample technol-ogy-transfer resources and its congressionaldelegation has often been able to channelfederal R&D funds to projects of demon-strable importance. No R&D project could beof greater potential value to the state’s mostbasic industry than the development ofenhanced safety technologies such as next-generation SCSRs. Without some such stimu-lus, it could be many years — even anothergeneration — before a next-generation SCSRis developed, tested for performance andreliability, approved by MSHA, and pur-chased by coal operators.

Miners working today do not have the luxuryof waiting that long for the safety protectionthey need now. It should be a high priority forWest Virginia, working with manufacturers,to take the lead in fast-tracking R&D anddeployment of the next generation of SCSRs.

56

4 Public hearing, Buckhannon, WV, May 2-4, 2006, transcript, p. 297.

5 John Nelson Boni, testimony, January 19, 2006, p. 60.

6 CSE Corporation, Instruction Manual for the SR-T Self-Contained Self-Rescuer Live Training Unit, p. 1.

7 CSE Corporation, Instruction Manual for the SR-T Self-Contained Self-Rescuer Live Training Unit,p. 12.

8Harley Joe Ryan, testimony, January 26, 2006, p. 52.

9 West Virginia Office of Miners’ Health, Safety & Training, “Memorandum to All West Virginia Under-ground Coal Mine Operators and Independent Contractors Who Perform Work in WV UndergroundCoal Mines,” July 10, 2006.

57

7 Why did the mine rescue effort take so long?

There is no end to the heartbreak.

Think about James Bennett, writing a mes-sage on a scrap of paper as he and elevenother miners awaited rescue behind a curtainserving as a makeshift barricade at the face ofthe Two Left section of the Sago Mine. It was4:25 on Monday afternoon, January 2, andthey had been trapped and waiting for helpfor almost exactly ten hours. The air was foulwith carbon monoxide. Three minutes earlier,at 4:22 p.m., carefully noting the time, he hadwritten: “Time is running out for us. We havenot heard anything from the outside yet.”Now he started another note: “I love you...”That was as far as he got.

4:25 p.m. At that hour, mine rescue teamswere still waiting outside the mine, stillwaiting for permission to go in. It would beanother hour, in fact, before the first teamswere allowed to begin exploring the mine —exploring from the portal, more than two anda half miles from where the Two Left crewwaited. And it would be more than another30 hours before rescuers reached the face ofTwo Left, No. 3 entry; heard the sound of aminer gasping for breath; tore back thecurtain being used as a barricade; and foundthe trapped miners. By that time — 41 hoursafter the explosion — all but Randal McCloy,

Jr., had succumbed to the carbon monoxidethat permeated the mine atmosphere.

For the miners’ families, and for the miners ofthe One Left crew who survived the blast,and for the mine superintendent and theother men who tried so courageously to go tothe rescue of the Two Left crew in the firstchaotic hours after the explosion — for all ofthem, the unending nightmare of the SagoMine disaster swirls around one centralquestion about the rescue effort: Why did ithave to take so long?

There is no satisfactory answer. The best wecan do in this report is to look at what hap-pened and try to understand, without un-fairly second-guessing anyone after the fact,what the thinking was in the rescue com-mand center. At this point in time, still await-ing the outcome of MSHA’s internal investi-gation, we have incomplete informationabout exactly how those decisions weremade. Based on what we know thus far,however, we need to consider why the rescueeffort largely failed and what must be done toimprove the odds that the next mine rescueoperation will be more successful.

The explosion occurred at 6:26 a.m. on Janu-ary 2. As soon as he became aware that the

58

Two Left crew had not been heard from andwas not responding to calls from mine dis-patcher Bill Chisolm, mine superintendentJeff Toler gathered maintenance chief DickWilfong, safety director Al Schoonover, andassistant maintenance chief Vern Hofer andheaded into the mine. They were in suchhaste that they did not think to bring a two-way radio or gas monitor with them. Theyencountered some miners not far inby theportal and told them to exit the mine. Then,pausing near a phone to check in with thedispatcher, they were heard by the escapingOne Left crew, who were making their wayout of the mine in the parallel escapewayentry. When a One Left crew member pokedhis head through a man door, hollering tomake the crew’s location known, Toler in-structed Wilfong and Hofer to take the menout of the mine on Toler’s mantrip.

The One Left foreman, Owen Jones, whosebrother Jesse was on the missing Two Leftcrew, refused to leave the mine. Jeff Toler,whose uncle, Martin Toler, Jr., was the fore-man of the Two Left crew, asked Owen Jonesto stay by a mine phone while he andSchoonover advanced inby on foot. Abouthalf an hour later, around 8:00 a.m., Wilfongand Hofer returned from the surface withradios, gas monitors, self-contained self-rescuers (SCSRs), and supplies to restoreblown-out ventilation. The five men thenadvanced further into the mine on themantrip. At 32 block they observed a blown-out ventilation stopping between the trackand the intake escapeway and hung curtainmaterial to keep the air flowing inby pastthem. They continued to advance in the track,hanging more curtain as they encounteredmore blown-out ventilation controls between32 block and 42 block.

At 42 block, their handheld gas monitorsalerted them to the presence of carbon mon-oxide (CO). They knew nothing about thecondition of the mine atmosphere ahead ofthem. Worried about triggering anothermethane ignition, they de-energized themantrip, disconnecting its batteries, and thenproceeded on foot in the intake air entries. Asthey advanced, they encountered thickeningsmoke and higher CO levels, and foundmultiple ventilation stoppings blown out.Stopping to continue hanging curtain toadvance the air flow, they had to wait for theCO levels to drop before moving ahead tohang the next curtain.

As they approached the outby end of themouth of the Two Left section, continuing tohang curtain, they realized that the CO levelswere taking longer to drop, and they foundthemselves getting into dense, swirlingsmoke. Thinking that the air flow must beshort-circuiting somewhere behind them,Toler asked Jones and Hofer to go outby,retrieve curtain, and look for ventilationcontrols in need of repair. Then he,Schoonover, and Wilfong made their way to58 block, at the mouth of the Two Left section.They could see nothing of the section fromtheir position near the intake escapeway: thesmoke was too thick.

Somewhere in there, they knew, the Two Leftminers were likely to be trapped — if theywere alive. A few times the three men heardsounds, and they yelled toward the sound,hoping it was someone coming toward them.But no one called back: most likely they werehearing distant falls of pieces of roof, jarredloose by the blast.

59

Toler, Schoonover, and Wilfong stood therefor awhile at 58 block, hesitating, trying todecide what to do next. They thought that ifthey tried to restore ventilation into thesection, they might be feeding fuel to a fire, orcreating an explosive mixture of methane andair. If they donned their SCSRs, which wouldgive them an hour of respirable air, theymight make it to somewhere near the face ofthe section, roughly half a mile from wherethey were standing, only to run out of air onthe way back.

Jeff Toler wanted to go in anyway. DickWilfong, who had been a mine rescue teamcaptain and was a veteran of fires and othermine emergencies, worried about the risks.

“I didn’t think we could go any further,”Wilfong later testified. “One of us would godown... I knew we was going to have to putour self-contained self-rescuers on. And wedidn’t have but a couple extras with us. Andthere was three of us there at that time... If weall put our rescuers on and tried to go in therewith no line or no ways to communicate otherthan hollering at each other, that we would —somebody would — maybe all of us wouldperish... And even if we went in there andfound somebody, what was we going to dothen? Say if we would have found six menor five men or four men. Then you’ve gottwo rescuers and you’ve got one on yourself,you know. Who do you give them to, youknow, if you’ve got more people than you’vegot rescuers? What do you do in a caselike that?”1

It was an agonizing decision. At about 9:30a.m., roughly three hours after the explosion,Jeff Toler picked up the mine phone at 58block and reported to the dispatcher that they

were abandoning their attempt to reach theTwo Left crew and were coming out —walking, because they did not want to takechances by re-energizing their mantrip. Toler,Schoonover, and Wilfong made their wayback into the intake escapeway, where,further outby, Jones and Hofer were waitingfor them. The five men would emerge fromthe mine at around 10:30 a.m.

By that time the mine had been placed undera so-called (K) order. And because there hasbeen controversy about what that meant, andhow the imposition of the (K) order influ-enced the subsequent mine rescue effort, weshould pause here to look at that question.

As described in the synopsis of events else-where in this report, International CoalGroup’s assistant safety director, JohnnyStemple, began trying to reach federal andstate authorities at about 7:30 a.m., roughlyone hour after the explosion. It was a holidayand he initially had no success, encounteringnothing but unanswered phones andvoicemail recordings.

At about 7:50 a.m., Stemple reached JohnCollins, an inspector for the West VirginiaOffice of Miners’ Health, Safety & Training(WVOMHST). Collins, who lives nearby inBuckhannon, arrived at the mine at about 8:15a.m. and issued a control order, which had theeffect of barring entry to the mine except formine rescue purposes (see the synopsis ofevents for the text of the order). He alsoissued instructions to begin monitoringcarbon monoxide, methane, and oxygenlevels in the air-return entry at the portal.(The mine had a blowing ventilation system,meaning that the mine fan forced air into theintake entry and that it then circulated

60

through the mine by means of an extensivenetwork of aircourses and ventilation controlsbefore exiting at the return entry.)

Then, at about 8:28 a.m., Stemple got throughto Jim Satterfield, an MSHA inspectionsupervisor, and described what he knewabout the situation at the mine: a possibleexplosion, two crews underground, one of thecrews unaccounted for, five men trying tomake their way toward Two Left. Satterfieldreportedly told Stemple he would contactother MSHA officials. At 8:32 a.m., he issued a103(k) order over the phone, instructingStemple (as Stemple recalls the conversation)that no one was to enter the mine or do anywork at the mine pending the arrival ofMSHA personnel and approval of a minerescue plan.2

Section 103(k) of the Federal Mine Safety &Health Act of 1977 reads as follows:

In the event of any accident occurring in a

coal or other mine, an authorized repre-

sentative of the Secretary [of Labor], when

present, may issue such orders as he

deems appropriate to insure the safety of

any person in the coal or other mine, and

the operator of such mine shall obtain the

approval of such representative, in consul-

tation with appropriate State representa-

tives, when feasible, of any plan to recover

any person in such mine or to recover the

coal or other mine or return affected areas

of such mine to normal.

A (K) order does not actually close a mine,nor does it mean that MSHA takes totalcontrol of the mine. The thinking behindSection 103(k) is that mine managementknows the mine better than anyone else andshould have the primary responsibility forputting a rescue plan into operation, but thatMSHA and state mine safety agencies such asWVOMHST have a statutory responsibility to

ensure that the operator’s plan is sound —and therefore, for the protection of miners,must have the authority to approve, disap-prove, or require modifications of the plan. Intheory, the imposition of a (K) order creates anorderly mechanism for shared decision-making — a way to expedite a well-plannedmine rescue operation while sharing responsi-bility for managing the risks involved. Overthe years it has often worked well.

In the aftermath of the Sago Mine disaster,however, there have been lingering concernsabout how the (K) order was implementedand whether it had the effect of impeding therescue operation. Satterfield was not physi-cally present when he issued the order, as theAct requires. Pending completion of MSHA’sinternal investigation, it is not clear whetherhe was fully informed of conditions at themine, including the fact that Toler’s initialrescue effort had gotten to within less than3,000 feet of where the trapped Two Leftminers were thought to be — and that therescuers had been able to proceed to thatpoint barefaced (that is, without having to usetheir SCSRs), an indication that much of themine atmosphere had not been severelyimpacted by the explosion and that theventilating system was largely functional andcould be quickly and easily repaired, at leastas far as 58 block.

This has given rise to the view, widely heldamong the victims’ families and by some ofthe other Sago miners, that MSHA, by issuinga (K) order, put the brakes on what otherwisemight have become an aggressive and per-haps successful rescue mission — and onethat could have saved precious hours bystarting in the vicinity of 58 block rather thanat the portal.

61

Given the long duration of the Sago Minerescue operation and its outcome, this view istotally understandable. But there are severalproblems with it.

First, the imposition of a (K) order does notabsolve anyone on the scene — MSHA,WVOMHST, or ICG — from responsibility fordeveloping, agreeing upon, and implement-ing the most aggressive rescue operation thatis deemed feasible under the circumstances.

Second, it is clear that the ICG personnel whowent underground in hopes of finding theTwo Left crew made their painful decision towithdraw based in part on their concernabout triggering a second methane ignitionand in part because neither they nor anyoneelse at the Sago Mine were equipped toexplore the potentially lethal mine atmo-sphere inby 58 block.

ICG had no one on-site who was trained andproperly equipped for such a mission. Al-though members of the One Left crew coura-geously volunteered to go back undergroundlater that morning to search for the Two Leftcrew, a strong case can be made that it wouldhave been criminally irresponsible to sendthem into a toxic, smoke-filled mine atmo-sphere equipped only with their SCSRs,unable to communicate to the surface frominby 58 block (because the mine phones inbyhad been knocked out by the explosion andtheir two-way radios had limited range),facing the possibility of encountering a fireand running the risk of triggering a secondmethane ignition.

Third, the evidence available to us thus farsuggests that at no time during the rescueoperation did ICG go to MSHA and

WVOMHST with a firm plan for a moreaggressive rescue strategy than the one thatappears to have evolved by some sort ofconsensus among those in the commandcenter. Had ICG done so, and had MSHA andWVOMHST refused to consider such a plan,there might be more of a basis for criticismthan now exists. While it may be reasonableto question whether the rescue strategyadopted by the command center was over-cautious, it is not correct to state that theimposition of the (K) order in and of itselfthwarted the rescue effort, either at 8:32 a.m.,when Satterfield imposed the order, orthereafter.

Like many mining companies, ICG had notrained mine rescue teams of its own at any ofits operations in West Virginia, Kentucky, andMaryland. The nearest ICG team was at itsViper mine in Williamsville, Illinois, morethan 600 miles away. Federal regulations ineffect on January 2 required every mine to beprovided emergency services by mine rescueteams located not more than two hours fromthe portal. For the Sago Mine, ICG had acontract with the Barbour County MineRescue Team, a volunteer association servingnumerous mines in the area.

Stemple tried to phone the Barbour Countyteam at about 8:00 a.m., only to discover,according to his subsequent testimony, thatthe team’s 24-hour answering machine hadbeen switched off.3 About 15 minutes later hesucceeded in reaching one of the members ofthe team, who then notified the team’s trainer,who called the others. After assembling theirgear, the first of two Barbour County teamsarrived between 10:40 and 11:00 a.m. (recol-lections and logs differ about the exact time),well over two hours after first being notified

62

and more than four hours after the explosion.About an hour later, around noon, the firstteam reported that it was ready to go into themine. But mine rescue protocol requireshaving a backup team available, and thesecond Barbour County team did not arriveuntil about 12:30 p.m.

Meanwhile, MSHA and state inspectors,assisted by some of the Sago miners, hadbeen taking gas readings at the return entry.The first readings, taken at around 8:30 a.m.,were not highly elevated: carbon monoxidewas recorded at 500 parts per million (ppm)and methane was measured at 1.1 percent ofthe mine atmosphere, well below the explo-sive range of 5 to 15 percent.

Toward noon, however, carbon monoxidelevels at the portal suddenly spiked upwardto 2,300 ppm, a development that increasedconcerns about the possibility of a fire in themine. The spike also created confusion in themine offices, where a rescue command centerhad just been set up, because the CO wasmaking its way from the mine pit into themine buildings. An MSHA inspector orderedthe buildings evacuated. Some of the miners’family members had gathered at the mineoffices, and they were escorted off the prop-erty (to the Sago Baptist Church, a few hun-dred yards away, where many of them wouldremain for most of the next two days). Then,as the evacuation proceeded, CO levelsdropped and the command center was

EFFECT OF CARBON MONOXIDE POISONINGAND COGNITION

Carbon monoxide (CO) poisoning and its relationship to adverse neurological symptoms is well estab-lished in the medical literature, and has been understood for decades by the public health and safetycommunity. When a person is overexposed to CO, their carboxyhemoglobin (COHb) level rises. Thehigher the CO level, the more severe the neurological symptoms will be, including headache, dizziness,weakness, nausea, vomiting, confusion, disorientation, and visual disturbances.1 This impairment to theneurological system can compromise a person’s judgment, and make it difficult process information andmake critical decisions. Because the brain is quite sensitive to the effects of CO,2 these symptoms canemerge quite rapidly after exposure, especially at high levels occurring after a coal mine explosion whenair flow is disrupted. It is reasonable to suspect that the miners trapped in the Sago mine had somedegree of CO-induced neurological impairment which may have affected their judgment with respect tosuccessfully finding an escape routes.3, 4 — Robert A. C. Cohen, MD

1 Raub JA, Mathieu-Nolf M, Hampson NB, Thom SR. Carbon monoxide poisoning:a public health perspective. Toxicology. 2000; 145: 1-4.

2 Kao LW, Nanagas KA. Carbon Monoxide Poisoning. Med Clin N Am. 2005; 89: 1161-1194.

3 Henz S, Maeder M. Prospective study of accidental carbon monoxide poisoning in 38 Swiss soldiers.Swiss Med Wkly. 2005; 135: 398-406.

4 Cohen RAC. Letter to J. Davitt McAteer, July 2006. Available at: www.wju.edu/sago/def.asp

63

reoccupied — only to be temporarily vacatedagain when the CO rose again, briefly, a fewminutes later.

By 1:00 p.m., CO levels at the mine mouthwere back down to 500 ppm, and methanewas measured at 1.1 percent. But the com-mand center considered the situation toovolatile to let the Barbour County mine rescueteam go in. Even after a more experiencedteam from Consol Energy’s Robinson RunMine arrived on site at around 1:45 p.m., theprevailing view in the command center seemsto have been that there was a strong possibil-ity of fire somewhere in the mine, and that tosend any mine rescue teams in withoutknowing more about constituent trends —that is, what was happening over time to themix of carbon monoxide, methane, othergases and oxygen in the mine atmosphere —would be irresponsible.

“We have a rich history of secondary explo-sions,” Ray McKinney, MSHA’s Administra-tor for Coal Mine Safety and Health, oftenpoints out, and he is right. One of the ex-amples that still haunts veteran mine rescuecoordinators is the Scotia Mine in LetcherCounty, Kentucky, where an explosion onMarch 9, 1976, killed 15 miners — nineinstantly, and six from carbon monoxideasphyxiation while awaiting rescue. Rescueteams recovered all of the victims withinabout 18 hours. Then, on March 11, while theroof was being rebolted to facilitate furtherexploration of the mine, a second and evenmore violent explosion occurred, killing 10miners and three federal inspectors. Rescueteams had to be withdrawn because of thefear of still another ignition, and the mine wassealed. The victims of the second blast could

not be recovered until the mine was reopenedabout a year later.

A more recent disaster served to underscoreMcKinney’s concern. On September 23, 2001,an explosion at the Jim Walter Resources No.5 Mine in Brookwood, Alabama, injured fourminers, including one who was too gravelyhurt to move. The other three injured minerswent to get help. The explosion had impairedthe mine’s ventilation system and methanewas rapidly accumulating, but 12 minersbravely tried to rush to the aid of the injuredminer. Fifty-five minutes after the first explo-sion, a second and more violent one occurred,killing 13 miners — the miner who had beeninjured in the first blast and those who hadtried to come to his rescue (one of whomsurvived the second blast but later died ofburns).

Because of nightmare scenarios such as Scotiaand Brookwood (and there have been others),McKinney and other senior MSHA officialsare understandably reluctant to send minerescue teams into harm’s way without know-ing as much as possible about the conditionsthey are likely to encounter. And on theafternoon of January 2, MSHA was at adouble disadvantage on that score.

MSHA owns two gas chromatographs —sophisticated instruments capable of measur-ing and analyzing multiple gases, includingthose most indicative of the presence of a fire.One is kept in Denver and the other in Pitts-burgh. Hampered by the holiday, MSHA wasslow in making arrangements to bring itsPittsburgh chromatograph to Upshur County,and it did not arrive at the mine until around5:00 p.m., about two hours after a chromato-graph owned by Consol arrived.4

64

Meanwhile MSHA was hindered in its abilityto interpret the gas readings at the mineportal not just by the absence of a chromato-graph but also because one of the agency’skey people, the man probably most able tointerpret gas chromatograph readings withassurance, was not at the mine — althoughthrough no fault of his own.

John Urosek, chief of the ventilation divisionof MSHA’s technical support group, was inColorado on January 2, monitoring a minefire that had been burning for weeks (fortu-nately without loss of life) in an Arch Coalmine. After learning of the situation at Sago,he was in the process of trying to obtain anairline ticket when Tim Martin, ICG’s corpo-rate safety director, happened to call ArchCoal, hoping to track down a firefightingconsultant, Ian Houlison. When Martinlearned that both Houlison and Urosek wereat the Colorado mine, he chartered a privateplane to bring them both to Buckhannon. Butit was close to 8:00 p.m. on January 2 —nearly 12 hours after the explosion — before astate police escort delivered them from theClarksburg airport to the mine site. It wasonly after that point that Urosek could begincontributing his interpretive expertise to themine rescue planning process.

Meanwhile, in late afternoon, Consol’schromatograph had begun augmenting thereadings from handheld gas monitors. At 5:25p.m., with CO readings at the portal declin-ing, Consol’s Robinson Run mine rescue teamwas finally authorized by the commandcenter to enter the mine, through the fanhousing, with instructions to explore the first1,000 feet inby the portal. Mine rescue teamshad been arriving at Sago throughout theafternoon, and the command center’s first

plan had called for one of them, Tri-State, tomake the initial exploration. But the RobinsonRun team was given the task becauseConsol’s mine rescue teams were the mostexperienced at the site.

At about the same time, a bulldozer went towork at the surface, building a rough road topermit drilling equipment to be moved to asite directly over Two Left. The commandcenter’s plan was to put a borehole down toassess the mine atmosphere in the area wherethe miners seemed most likely to be trappedif they had survived the initial blast.

Both projects — exploration of the mine andconstruction of the borehole — proceededslowly as the evening wore on. On the sur-face, there were delays while surveyors usinghigh-grade Global Positioning System receiv-ers tried to determine precisely where to sitethe drill rig. Because of unfavorable weatherconditions they were having trouble gettingaccurate readings from orbiting GPS satellites.Exploration underground, meanwhile, wasslowed by the command center’s instructionsto the mine rescue teams to work systemati-cally back and forth across the multipleentries of the mine as they advanced, takingmultiple gas readings and thoroughly check-ing conditions before proceeding further.

That cautious and deliberate approach seemsto have been intended primarily to ensurethat the rescuers would not inadvertentlyproceed past a possible ignition source, and itmade sense in the context of MSHA’s fear ofsecondary ignitions. There is little doubt thatthe “rich history” was much on the minds ofthe MSHA officials in the command center,some of whom had been involved in theagency’s investigation of the Brookwood

65

disaster. No one had to tell them to worryabout the safety of the mine rescue teamsexploring the Sago Mine.

On the other hand, there had to have beensome concerns that the operation was pro-ceeding too slowly. Everyone involved inmine rescue knows that time is the enemywhen coal miners are missing after an explo-sion or in a fire. With every passing hour, theodds of finding anyone alive rapidly dimin-ished. The command center must have beenacutely aware of that, just as the mine rescueteam members certainly were.

The dilemma, to be understood, must be seenas something like war. It is a maxim thatgenerals tend to want to fight the last war,rather than the one in front of them, becausethe last war is the one they know. So they maybe reluctant to try new strategies, even whenthe situation calls for fresh thinking. Mineemergencies pose a similar kind of challenge.Senior MSHA officials know all about Scotia,Brookwood, and other cases where thingshave gone terribly wrong. That awareness isdrilled into their training and inevitablycolors their thinking. Nobody wants to beremembered for having lost a mine rescueteam. But history teaches some harsh lessons.Lincoln had to replace his cautious generalswith men like Grant and Sherman, who areremembered today not for the losses theytook but for getting the job done. The last-waranalogy should not be pushed too far, but itshould not be ignored.

As Mark Chewning, one of the captains of theBarbour County Mine Rescue Team, later toldinvestigators: “You know, I’ve been doing thisfor, I think, 21 years now, and they should letthe mine rescue people have a little more

leeway in what to do and what not to do... Imean, since we’re the ones being trained todo this, we should have a say in it somewhere— you know, say ‘Hey, you know, we’ll try it.’We know the danger to begin with or wewouldn’t be in it... Like, these 12 guys in there— if we was willing to take the chance, to gotry, I think we should have had the option togo try, if we sign a waiver or however youwant to do it.”5

That view did not prevail in the commandcenter. One result was that problems, as theyarose, tended to have the effect of slowing therescue operation or stopping it cold. Forexample, when water was encountered in thereturn entry, considerable time was spentrestoring power to a pump, starting it, andwaiting to let rescuers advance until thepump brought the water level down. Simi-larly, when rescuers spotted a red light near34 block, and determined that it was a COmonitoring station that should have been de-energized when power was cut to the mineafter the explosion, all of the rescue teamswere pulled out of the mine while the prob-lem (a battery backup power source that hadto be disabled from the dispatcher’s com-puter) was dealt with. While actions such asthese were definitely prudent, they alsoadded time to the rescue operation when timewas already at a premium.

On the surface, meanwhile, the drilling of theborehole got under way at about 2:45 a.m. onJanuary 3, approximately 20 hours after theexplosion. The command center, after pullingthe rescue teams out of the mine, decided tokeep them outside until the borehole hadpunched through into the Two Left section,because of concern that the drill might ignitemethane. The drill punched through at about

66

5:30 a.m. Work stopped and the drillingequipment was shut down a few minuteslater in order to listen for sounds from themine. There were none. The drill steel wasthen struck in the hope that someone in themine would hear the sound and send a signalback to the surface. No signal came. At about6:30 a.m., while a video camera was beinglowered into the borehole, rescue teams weresent back underground. The camera showedno signs of physical damage from an explo-sion, but it also showed no signs of the TwoLeft crew.

More than 24 hours had now passed since theexplosion. The story of the second day isessentially one of continuing slow movement,punctuated by setbacks. For example, MSHAhad brought a robot to the mine, and sent it inat around 7:30 a.m. on January 3 with thehope that it could accelerate the rescueoperation by monitoring the mine atmo-sphere ahead of the teams. The robot was alsoequipped with a camera, and theoreticallycould travel up to 4,000 feet by remote control— meaning that it would be able to reach themouth of the Two Left section from its start-ing point at 31 block. Instead, the robotfunctioned for only about an hour beforegetting stuck and becoming disabled on amuddy and uneven patch of the mine floor,where it stayed for the duration of the rescueand recovery operation.

Frustration — with the robot, with the pace ofthe rescue — seems to have been building inand around the command center. Later, ICGcorporate safety director Tim Martin de-scribed it to investigators: “There was somehigh emotion there, obviously, that we have....13 men unaccounted for and we need to getto those folks... [yet] they were actually using

a method almost like a mine rescue contest totunnel across and back and just being ex-tremely methodical about exploring andmeasuring the constituents.”6

According to Martin, who was near but not inthe command center, he huddled withMSHA’s John Urosek and firefighting consult-ant Ian Houlison to talk about accelerating themine exploration. They agreed, Martin said,that it would be feasible to advance morequickly by limiting readings to fewer entriescloser to the intake. They discussed whoshould present this idea to the commandcenter, and agreed that Urosek should do it,Martin recalled: “And it took about tenminutes but they finally accepted the idea,and that’s when the teams kind of picked uppace and started moving.”7

At about 2:00 p.m., rescue teams reached themouth of the One Left section. At the com-mand center, meanwhile, a decision wasmade to bypass exploration of One Left, otherthan the openings at the mouth of the section,since it was reasonably certain that there wasno one there (the One Left crew had comeout, and the assumption was that firebossTerry Helms had left the section, on his wayto Two Left, before the explosion occurred,because it was known that his lunch buckethad been left for him at the Two Left switch).About 15 minutes later, at 50 block, the teamsfound the One Left mantrip where the crewhad abandoned it, with its headlights still on.

At about 4:20 p.m., the command centerdecided to have the rescue teams move theirfresh air base up to 57 block. While some ofthe team members were hanging curtainoutby the base site, to direct the air flow overit, others explored the entries between 57 and

67

58 block. At about 5:15 p.m., MSHA rescueteam member Ron Hixson, working his wayforward alone in the track entry on his way tojoin a Consol team as a backup man, cameupon the soot-covered body of Terry Helms,lying across a rail near the coal rib. (A subse-quent examination indicated that Mr. Helmshad probably died instantly, after beingcaught in the direct path of the explosion.)

Shortly after 5:45 p.m., with the new fresh airbase established, rescuers advanced towardthe sealed area to check the mine atmospherethere. Not knowing that the seals had beenalmost completely destroyed, they actuallywalked through the pulverized rubble with-out realizing they had done so until theyreported their location — and some very highCO readings — to the command center. Theythen retreated, working their way across theentire set of seals and confirming that theseals had been blown out by an explosionoriginating in the sealed area.

At this point — some 35 hours after theexplosion — the command center finally hada clear sense of what had happened on themorning of January 2. Until this point, or atleast until the borehole camera revealed theabsence of damage on the Two Left section,there had been lingering speculation that theexplosion might have originated near the faceof the section, perhaps triggered by the TwoLeft crew, arriving and energizing equipmentin the presence of an explosive methane-airmix that they had not detected. Now, with therealization that the crew most likely had notbeen killed outright by the explosion, therewas a heightened sense of urgency aboutfinding them.

Advancing into the Two Left section, rescueteams found all of the ventilation controls inthe primary escapeway damaged from themouth of the section to crosscut 12, more thana third of the way toward the face. Then, ataround 7:30 p.m., rescuers spotted the TwoLeft mantrip in the track entry at crosscut 10.Searching the area, rescuers found no one butreported to the command center that itappeared as though the crew had tried to usethe mantrip to escape and had been forced toabandon the vehicle after encountering debrisfrom the explosion across the tracks. Rescueteam members found footprints in the intakeescapeway, indicating that the Two Left crewhad walked there from the abandonedmantrip, but the footprints disappeared —possibly lost in the mud and water — so therewas no indication of how far the crew mighthave gone before turning back. (The com-mand center was also concerned about thepossibility that they had not turned back —that they had gotten disoriented in the smokeand stumbled past the destroyed seals intothe abandoned area — but the rescue teamsreported having seen no footprints there.)

Just after 8:00 p.m., rescuers exploring theintake escapeway found the top and bottomcovers of 12 SCSRs at crosscut 11. That in itselfdid not indicate where the men were, butlogic suggested that if they had been unableto make their way out of the mine in themantrip, they most likely would have re-treated in the other direction, toward the face.But it was also possible that they could havebarricaded themselves somewhere else in thenearly half-mile-long section. The rescueteams, switching out according to protocol sothat no team would become exhausted,needed nearly three hours to explore theentries from the mouth of the section tocrosscut 15, just over halfway to the face.

68

At some point between 11:00 and 11:30 p.m.(the exact time is not shown in the commandcenter’s logs), the rescue teams were givenpermission to break with protocol and make apush to the face of the section. Normal minerescue protocol would have required them toadvance no more than 1,000 feet beforeestablishing another fresh air base and hang-ing curtain to direct air to the new base. Thatprocess could have added another two hoursor more to the exploration of Two Left. AsRon Hixson later testified: “Yes, we were

going to go further than 1,000 feet. The airquality was bad, but it wasn’t a situationwhere you would die or anything if yourapparatus failed or if you had a major prob-lem. We felt the risk was worth taking and wehad the radios to communicate back andforth, and we went for the face.”8

But to push straight for the face meant thatcommunication would be difficult. The minephone system could be used to communicatebetween the command center and the fresh

Handheld and hardlineat 59 block

Handheldat 9 block

Handheld atpower center

Locationof miners

Hardline phone at 58 blockcommunication withcommand center

One L

eft

Tw

o L

eft

Sealed Area

Communications were stretched to the limit at the moment when mine rescuers found thetrapped miners in the No. 3 entry of Two Left. A rescuer under apparatus equipped with ahandheld two-way radio in the track entry near the power center had to relay messages toanother rescuer under apparatus with a handheld near crosscut 9. He then relayed messagesto rescuers in respirable air at 59 block. They communicated by cable to rescuers at the freshair base at 58 block, who communicated by mine phone to the command center at thesurface. (Light grid at upper right shows coal blocks not yet mined.)

69

air base at 58 block. Moving inby from thefresh air base, a couple of hundred feet intothe neck of the Two Left section, rescuersequipped with a cable reel strung a cable thatallowed them to stay in touch via headsetwith the rescuers manning the fresh air base.Because all of these mine rescue team mem-bers were working in respirable air, they werenot wearing breathing apparatus, and so theirhardline communications were reasonablyclear and reliable.

But further into the section, where the COlevels were elevated, rescuers had to be underapparatus, communicating to each otherthrough voice diaphragms in their facemasks. Some also had handheld two-wayradios, which when fully charged had aneffective range of about 500 feet — as long asthey were in direct line-of-sight with eachother. That was feasible if two men stayedwithin 500 feet of each other in the sameentry, but if one of the men was moving backand forth between entries, the other man wasunlikely to be able to hear him clearly, if at all.

To help deal with the problem of stretched-out communications, one of the rescuersequipped with a two-way radio stayed in thetrack entry at crosscut 9, roaming a few yardsinby or outby as necessary to pick up trans-missions. He was theoretically able to relaymessages from inby to the rescuers at thecable reel, who could then relay them by cableto the rescuers at the fresh air base, who couldthen relay them by mine phone to the com-mand center. But this four-step communica-tion linkup was inherently fragile, as wouldsoon be demonstrated.

The push to the face was led by a team fromConsol’s McElroy Mine, accompanied by one

federal and one state inspector — MSHA’sRon Hixson and WVOMHST’s Bill Tucker.The rescuers were grateful for the order tobreak with protocol. (As McElroy teamcaptain Jimmy Klug said later: “By the way,whoever made that decision was a goodperson to do that.”9) They advanced rapidlyin the track entry, knowing that the odds offinding anyone alive after so many hourswere very slim — but carrying a dozen extraSCSRs with them, just in case.

At about 11:40 p.m., they reached the face,and split off, with Klug and Tucker exploringthe entries to the left, Hixson and McElroyteam members Mike Clark and Jim Smithexploring to the right. Making their way intothe last open crosscut just outby the No. 3entry, Klug and Tucker suddenly heard adeep gasping sound at about the same mo-ment that they caught sight of a curtain hungdiagonally across the entry.

As Klug pushed past the curtain, the beam ofhis cap lamp showed him that he had foundthe missing miners. Some were sitting againstthe coal rib. Some were lying down. All weresilent and motionless, except for the minerwho was the furthest inby, closest to the face.Klug heard him gasp again and ran straightto him.

The miner was in a sitting position. Anotherminer sitting next to him, a bigger man, hadtoppled across him. Klug struggled to pull thegasping miner free. Bill Tucker, hurrying intothe entry just a step or two behind Klug,turned back into the crosscut momentarily toyell for help.

As Tucker later testified to investigators: “Iwas screaming, ‘We need help! We need help!’

70

And I mean, when you first looked in, in mymind at that point, you know, I thought thatmost of them were dead. But I was holleringback for help and I said, ‘They’re over here!They’re over here and they’re alive!’”10

No one should ever blame Bill Tucker formiscommunicating the situation. At least oneminer was alive, and in the instant of discov-ery there had not been time yet to check all ofthe others. However, as Hixson, Smith, andClark responded to Tucker’s call and rushedinto the entry, it became obvious to them thatthe other miners were not alive. Klug latertestified: “We all knew. I mean, I don’t wantto — we knew.”11

But they checked anyway. While Klug andsome of the others moved the gasping mineronto his back, fearful that he was about toexpire and hollering at him in a desperateeffort to bring him back to consciousness,rescuers went from miner to miner, shakingthem, calling out to them, feeling for pulses.The other miners were gone, all of them. Forthe rescuers there was absolutely no doubtabout it.12

But that was not the message that made itsway through the four strained links of thecommunication chain to the command center.It is still not clear how the word reached therescuer with the two-way radio in the trackentry at crosscut 9, but what he remembers isthat someone transmitted from the face area:“We need help, we’ve found them, we foundall the men, we need help... We need medicalhelp. We’ve got two people we’ve got down,we’ve got to have stretchers, we need help.”13

Hixson remembers how the confusion aroseabout a possible second survivor. As more

rescuers rushed forward to the face, respond-ing to the first call for help, they too checkedthe lifeless miners, shaking them and callingout to them. One of the rescuers pulled one ofthe victims away from the rib, and when hedid, as Hixson recalls, “I think it was just airtrapped in his lungs, but he let out a loadmoan, and we thought we had a second manalive.”14 It was only a matter of momentsbefore the rescuers realized their error, but themessage went back before they could cancelit.

Chris Lilly, captain of the Tri-State MineRescue Team, was at the fresh air base, com-municating with the command center via thehardline mine phone, when word came fromthe next inby link of the communication chain— the rescuer at the cable reel, who in turnwas relaying messages from the rescuer withthe two-way radio at crosscut 9. Lilly remem-bers the message: “It may have been, ‘Weneed supplies, oxygen, stretchers, 12 alive.’ Orit was ‘Twelve alive, we need oxygen andstretchers.’... It was either supplies first orwhat they found first.”15 Whichever it was,Lilly remembers asking for confirmation:“They confirmed it. And then we reported itoutside to the control center.”16

(Memories conflict on the question of confir-mation. MSHA rescuer Frank Thomas wasalso at the fresh air base when the messagecame from the rescuer at the cable reel. Helater testified: “The guy on the hard line[cable], the words came out of his mouth,‘Twelve alive.’ The guy on the mine phonecalled out [to the command center], ‘Twelvealive.’ Nothing was confirmed before hecalled outside.”17)

71

Although the situation at the face was chaotic,the evidence suggests that very little timeelapsed before the rescuers got on their radiosagain to report what they all quickly under-stood — that there was just one survivor.Hixson, realizing they were going to needmore people to help carry the survivor, got onhis radio and called out: “Hey, we have all 12guys accounted for. We have one alive, andwe need help, and we need help now.”18

The rescuer with the relay radio at crosscut 9,running back and forth as he tried to maintainreception from inby, heard that message. Inorder to transmit it, he had to run a few yardsoutby in the entry. He remembers transmit-ting: “They’ve only got one person alive.”19

Ordinarily he would have waited for aresponse, but this time he didn’t — becausehe saw lights coming, realized it was therescuers carrying the survivor, and “I just ranup to meet them, to help any way I couldhelp.”20

They needed help. Carrying anyone on astretcher in a coal mine, stumbling overterrain illuminated only by cap lamps, istough work. Carrying the sole survivor of thedisaster was made even tougher by the factthat he was unconscious, perilously close todeath, and in desperate need of oxygen. Andthe stretcher-bearers had to make their waythrough the rubble from the explosion inaddition to stumbling through mud andwater that was sometimes up to their knees.

In the entry, upon finding the survivor, Klughad broken open an SCSR and forced themouthpiece between the survivor’s lips, but itwas hard to keep the mouthpiece in place andeven harder to get him to breathe through it.And once the survivor was on a stretcher and

being carried back toward the fresh air base,making the SCSR work for him became amajor challenge. It meant running alongsidethe stretcher, holding the mouthpiece in thesurvivor’s mouth, holding the SCSR along-side him, peering at him constantly to see ifhe was breathing, avoiding falling towardhim or away from him, and trying not to gofaster or slower than the stretcher bearers —who were still under apparatus, their facemasks fogging over with sweat, tiring rapidlyas they stumbled and slipped. It was a night-marish carry. If a man felt about to collapse,he would yell out and another rescuer wouldslide in to take his place, taking care not tolet the stretcher drop, struggling to maintainthe pace.

It was at about 12:15 that the rescuers bearingRandal McCloy, Jr. — they didn’t know hisidentity yet — reached fresh air at the neck ofthe Two Left section. By that point they hadused three SCSRs in order to keep giving himthe initial oxygen shot supplied by the smallbottle within each CSE SR-100.21 Now, withthe rescuers able to proceed barefaced inrespirable air, they borrowed a rescuer’sDraeger BG-4 breathing apparatus and fittedit to McCloy’s face. The BG-4 is a positivepressure device, meaning that it suppliesoxygen whether the user works at breathingor not. Soon the rescuers noticed that McCloyseemed to be breathing better — not normallyby any means, but no longer taking a breathonly every five seconds or so. Encouraged,they managed to keep moving rapidly as theycarried him back to a mantrip at the mouth ofthe One Left section, and then, in the mantrip,headed out of the mine.

On their way out, they encountered anunexpected sight, coming toward them —

72

a motor pulling a supply car crowded withpeople. There was confusion as the rescuershollered to clear the track so they could keepgoing with the survivor.22 The incomingminers — including mine superintendent JeffToler — wanted to know where the othersurvivors were. As Frank Thomas remembersthe moment: “Someone gets out of themantrip that was coming in and asked,‘Where’s the rest of them?’ And in our bunch:‘The rest of who?’ ‘The rest of the men, theother miners.’ ‘They’re all dead.’ Completesilence.”23

The confusion had ramped up at 11:45, whenthe command center got the message from thefresh air base that all 12 miners were alive. TyColeman, International Coal Group’s safetymanager for West Virginia and Maryland,was ICG’s man in charge in the commandcenter throughout the rescue operation. As helater testified: “I was just basically listening tothe mine phone and it came over the[speaker] phone… ‘We found 12 alive.’ Andthis was late in the hour. And I thought, youknow, maybe I didn’t hear right. And theroom got loud and I [can’t] remember whowas next to me, and I said, ‘What did he say?Did you hear what I [heard]?’ He said, ‘Yeah,12 alive.’…It was absolutely euphoric, thefeeling there… But unfortunately I said, ‘Youknow, we got to confirm, you know, let’sclarify it.’ And the room erupted, but we werestill doing business. And I had never seen somany old hairy guys cry in my life.”24

There have been many accounts of the roomerupting in celebration. And after all thoseanxious hours, it was only natural to cel-ebrate. There is no evidence, however, thatColeman or anyone else in the room thoughtto request confirmation, despite their aware-

ness of the tenuous communication chaininside the mine and what has to have beentheir acute awareness of the odds against allof the miners surviving in bad air for such along time. They did not follow basic emer-gency protocol requiring verification of allcommunications. Instead, reports of themiraculous discovery — “twelve alive!” —almost instantly reached the miners’ familieswaiting at the Sago Baptist Church.

Worse yet, it appears that one or more ICGemployees may have taken it upon them-selves to provide details to the families thatsimply could not have originated with therescuers underground. For example, severalfamily members vividly remember a com-pany official telling them that as soon as theminers emerged from the mine they weregoing to come over to the church before goingto get medical treatment.25

That struck some of the family members aswildly implausible. As Amber Helms, daugh-ter of Terry Helms, said: “Any man who istrapped underground for that long a time isnot going to have a choice of whether theywant to walk to a church. They’re going to beput in an ambulance and they’re going to takethem immediately [to a hospital].”26 Butwhatever doubts the families may have hadwere pushed aside by the apparent certaintyof the reports from the mine and the sheer joyof hearing that their loved ones were alive.

No one cautioned them to wait for verifi-cation or further bulletins. Moreover, itappears that no one from MSHA orWVOMHST briefed the families at anytime during the entire rescue operation —an appalling failure.

73

Meanwhile, at the command center, prepara-tions got under way to send medical suppliesinto the mine and to summon ambulances totake the miners to the hospital. It was notuntil almost three-quarters of an hour later, at12:25 a.m., that clarification came — to thecommand center, not the families — and eventhen the command center failed to under-stand at first.

When the rescuers carrying Randal McCloyreached the neck of the Two Left section, neartheir fresh air base, they had been met byother rescuers hurrying inby on the assump-tion that all 12 miners were alive. Realizingthat the original miscommunication hadtaken on a life of its own, Jimmy Klug went tothe mine phone at the fresh air base andcalled outside to the command center.

Earlier, before going underground, Klug andsome of the other rescuers had been told touse a code to communicate with the com-mand center in the event that they foundanyone, because some of the people in thecommand center had become concerned thatthe news media were somehow eavesdrop-ping on their communications. The code for“body” was “item,” and each of the missingminers was to be identified by a number. Therescuers had used the code when they foundTerry Helms, reporting to the commandcenter that they had found one “item” in thetrack entry. Now Klug used the code again,after realizing that the command centerapparently did not know that all but one ofthe miners were dead. He recalls saying overthe mine phone: “We have 11 items.”27

The message didn’t seem to sink in, accordingto Tim Martin, ICG’s corporate safety director,who heard what Klug was saying. Martin had

had a role in coming up with the code. Heremembers that the command center at thatmoment was “kind of full and the emotionswere very high that they thought they had 12miners alive. The word came across from[Klug] on the phone that ‘We have one on astretcher and 11 items.’ And it just hit me likea ton of bricks that I knew what that meant.And I kind of stepped back for a minute — Ijust, you know, was in shock. But then I couldsee the rest of the room didn’t understand.”28

Klug, as captain of a Consol team, was speak-ing with a Consol safety director, RickMarlow. Martin remembers saying to Marlowand the others in the command center: “Ithink we got bad news here... You need toconfirm that. And they argued with me for ashort period of time that, ‘No, no, no... we got12 people coming out’... And I had to kind ofraise my voice pretty loud to get them tolisten to me. I said, ‘Listen, he... just said wegot 11 items.’ I said, ‘Have him repeat whatwe have as far as survivors.’... And theanswer came back, ‘We have one on astretcher and 11 items.’ And they turned tome and said, ‘See, we’re okay,’ or somethingto that effect. I said, ‘Listen, are you all forget-ting about the code?’ I said, ‘Tell them to dropthe code and just tell us what they’ve got asfar as survivors.’ So they called back in andsaid, ‘Do you have 11 survivors and one on astretcher?’ And they said, ‘No, it’s the otherway around.’”29

That message staggered the command center.Klug recalls Marlow telling him: “‘Forget thecode. What do you mean?’ I said, ‘There’s 11deceased people.’ And he just — he wasspeechless.”30 Klug remembers that Marlowthen told him: “Just bring everybody outside,we’re going to regroup and go back in.”31

74

The rest of the night’s tragedy can be quicklytold, in part because it dominated the nextday’s news stories and in part because it doesnot directly bear on this narrative of the minerescue operation, which in reality came to anend at 11:45 on Tuesday night, January 3, withthe discovery of the trapped miners and therealization that all but one were deceased.

Randal McCloy’s rescuers brought him out ofthe mine at about 1:00 a.m. on Wednesdaymorning, January 4, after being briefly de-layed when they encountered the incomingpersonnel, including a doctor — who, sincethey had no communication with the com-mand center, were still under the impression,up to that moment, that they would befinding 12 miners alive. Mr. McCloy’s re-moval from the mine marked the conclusionof the rescue effort — approximately 41 hoursand 15 minutes after the Two Left miners firstfound themselves trapped by the smoke anddebris from the explosion in the sealed area ofthe mine.

There are three lingering questions about therescue operation that must be addressed:

• Why did the command center, after learn-ing about the miscommunication from withinthe mine, take so long to communicate thetruth to the miners’ waiting families?

• Why did the command center have somuch difficulty determining the facts in thefirst place?

• And, finally, what can be done to improvethe odds that the next mine rescue operation— always a race against time — will notrequire nearly two full days to complete?

As previously noted, it was at about 12:30a.m. on Wednesday when McElroy teamcaptain Jimmy Klug succeeded in communi-cating to the command center that the rescu-ers were bringing out just one survivor on astretcher and that the rest of the Two Leftminers had died. The command center’sresponse to that was to order everyone out ofthe mine — even though the mine rescueteam from ICG’s Viper Mine was under-ground at the time, assisting in the removal ofRandal McCloy. Several members of the Viperteam were certified emergency medicaltechnicians (EMTs), fully capable of reassess-ing the condition of the other Two Left min-ers. But they were ordered out of the minealong with everyone else.32

ICG president and CEO Bennett Hatfield,who was in the command center at the time,subsequently testified that the commandcenter was informed that there “appeared” tobe only one survivor and the other miners“appeared” to be deceased33 — a statementnot supported by the testimony of Klug andother rescuers, who do not appear to havehad, or communicated, any ambivalenceabout the status of the miners. Nevertheless,Hatfield also testified: “The immediatereaction in the command center was that thisreport of only one survivor may be erroneous.Many participants in the command centerclung to the fervent hope that the other 11miners may be in some sort of comatosestate.”34 If indeed they were clinging to sucha hope, it is all the more difficult to under-stand why they ordered everybody out of themine rather than immediately sending a teamback to the face of Two Left.

The command center seems to have been in astate of general confusion at that point —

75

understandable given the level of fatigue androiled emotions that everyone there musthave been struggling with, but it continued toexperience problems. For example, whenrescuers brought the survivor to the surface atabout 1:00 a.m., no one managed to determinehis identity before sending him in an ambu-lance to St. Joseph Hospital in Buckhannon.It was not until more than an hour later thatthe command center learned which minerhad survived, and in the meantime there wasno information that could be transmitted tothe families about his identity.

ICG’s Viper team, exiting the mine with theother rescuers shortly after 1:00 a.m. onWednesday, January 4, walked up from thepit to the command center. Team captain PeteBryant later testified that no one debriefedthem about what they and the other rescuershad seen in the mine.35 However, when theyencountered ICG corporate safety directorTim Martin, he told them, according toBryant, to “get back in there and see if therewere any folks still alive.”36 Someone thensupplied the team with five stethoscopes andsome first aid equipment, and they headedback into the mine — perhaps 20 minutesafter exiting it, as Bryant recalls.37

There is some confusion on that point. Ac-cording to MSHA’s log, all rescue personnelwere out of the mine at 1:20 a.m. The log alsoshows that the command center, after decid-ing to send the Viper team back undergroundto reassess the condition of the Two Leftminers and remove them from the mine,supplied the team with body bags as well asstethoscopes. The rescuers then re-entered themine at 1:55 a.m, according to MSHA’s log.At 2:35 a.m., they reported to the commandcenter that they had reassessed the minersand that all were indeed deceased.38

At that point — nearly three hours afterrescuers had first determined, beyond anydoubt in any of their minds, that only one ofthe miners in Two Left was alive, and morethan two hours after the initial miscommuni-cation had been cleared up — Hatfield andother ICG officials went to the Sago BaptistChurch to deliver the horrific facts to theminers’ families.

Hatfield subsequently testified that an at-tempt had been made at about 1:45 a.m. tosend a message to the church that there were“conflicting reports” on survivors and thatthe initial reports might have been “toooptimistic.”39 But by 1:45 a.m. there were no“conflicting reports” — none at all.

Subsequent interviews with the rescuers byfederal and state investigators confirm thatnone of them, as they exited the mine, hadany doubt whatsoever about the condition ofthe miners they had left behind — nor is itremotely imaginable that any of them wouldhave been willing to leave the Two Leftsection if they had believed that any of theminers there were still alive. The only confu-sion on this point seems to have been in thecommand center. In any case, the messagesent at 1:45 a.m., although reportedly relayedto the church by state police officers, neverreached the families.40

ICG officials have offered various explana-tions for the confusion in the commandcenter, the failure to confirm informationcoming from underground, and the failure tohold that information within the confines ofthe command center. “Those of us involved inthe rescue effort suffered from the samefatigue and range of emotions that everyonein the church did,” Hatfield later testified.41

76

Ty Coleman, in charge of the rescue operationfor ICG, testified that he was in the commandcenter for the entire 40-plus hours of therescue operation and “did not relinquishcommand responsibility” at any time fromthe morning of January 2 until the bodies ofthe miners were brought out of the mine onthe morning of January 4.42

That is a long stretch of time. Fatigue maywell have played a part in the commandcenter’s confusion, and it should be notedthat — as MSHA is well aware — emergencymanagement experts routinely cautionagainst letting managers stay on station toolong, to the point where sleeplessness andexhaustion can impair judgment and deci-

77

sion-making. No one should question thededication of anyone in the command center,but if we are to learn anything from the Sagodisaster it should be fair to point out thatfatigue must be managed just like every otherfacet of an emergency.

As to the question of why the commandcenter failed to control dissemination of theinitial misinformation about finding 12miners alive, Hatfield blamed the lapse on“cellphone communications,”43 and ICG vicepresident Charles Snavely testified: “With allthe people coming in and out of the com-mand center, evidently people overheard thiscommunication and reported that to thefamilies at the church.”44

Cellphones unquestionably played a role inthe miscommunication to the families, and itshould be noted that Sago was the first mineemergency to occur in this age of ubiquitouscellphones. Their omnipresence strengthensthe case for ensuring that firm control ofcommunications, preferably through pre-designated spokespersons, becomes part ofany emergency management plan in thefuture.

On the other hand, throughout the rescueoperation at Sago there appears to have beenlittle if any control over who came and wentat the command center, and little effort seemsto have been made to keep critical informa-tion within its walls. At the moment when thecommand center heard that all 12 miners inTwo Left were alive, ICG’s Tim Martin wasoutside, on his way back to the commandcenter from the mine’s bathhouse, where hehad been helping a mine rescue team stage itsequipment in preparation for going into themine. As he later testified, he heard “several

people” out in the open commenting on the“great news.”45 Someone’s cellphone mayhave been used to get the news to the fami-lies, but that could not have happened if thenews had been kept confidential until it wasconfirmed. The information could not havebeen “overheard” because no attempt wasbeing made to guard it against being over-heard. MSHA, the state, and the companyshare the responsibility for that failure.

MSHA rescuer Ron Hixson probably spokefor all of the rescuers when, while beinginterviewed later by federal and state investi-gators, he was asked if he had anything hewanted to add.

“The only thing I’d like to say is that... I knowthere was a lot of press and there was a lotwent on with the communications,” Hixsonsaid. “And as far as the guys underground,speaking for myself on down the line, I’m notblaming anybody... but we were in a verydifficult situation. If that communicationswould have stayed where it was supposed to,at the command center, and not got out, noneof that would have happened. And I feel asbad for those families — I mean, that was aterrible thing that they had to go through. Ijust think as far as mine rescue, you know,any time I’ve ever been involved in it, whenwe’re calling outside, that information ischecked and double-checked and triple-checked before it’s ever written down, beforeit’s ever — you know, it’s verified. And itnever had a chance to be verified this time.”46

MSHA’s internal investigation, not yet com-pleted as of this writing, may address some ofthe not yet answered questions about how thecommand center functioned. In the mean-time, it is not our purpose here to question the

78

commitment of any of the individuals —federal, state, or corporate — who had thedemanding and ultimately heartbreaking task of trying to organize a race against time.But there are questions, of course, about theinitial delays in getting the mine rescue effortlaunched. And there is a larger question thatwill always hang over the Sago rescueoperation:

Could precious time have been saved bystarting the rescue at — or at least close to —58 block, given that the mine outby that pointhad been at least partially explored, the air tothat point was respirable, and gas measure-

ments were not showing exceptionally highor explosive elevations?

That is a hard question to answer with anycertainty. Those conditions were known as of9:30 a.m. on the morning of the explosion,when Jeff Toler, Dick Wilfong, and AlSchoonover were still at 58 block. But then,rightly apprehensive about a possible secondexplosion and further loss of life, they exitedthe mine. Once they were outside, however,no eyewitness accounts of conditions inby theportal were available to the command center.What if they had stayed at 58 block instead ofexiting the mine? If they had been able to

On Thursday, January 19, 2006, just 17 days after the Sago Mine disaster, a fire broke out in MasseyEnergy’s Aracoma Alma No. 1 Mine in Melville, Logan County, West Virginia. Twenty-one minerswere inside the mine when a carbon monoxide detector alongside a conveyor belt began sounding analarm at about 5:35 p.m. Nineteen miners made it out of the mine safely. Two miners — Don Bragg andElvis Hatfield — became separated from a crew of 10 others and did not escape.

Mine rescue teams entered the Alma mine at around 11:30 p.m., about six hours after the belt fire wasdiscovered and while it was still burning. Rescue teams using water and foam suppressed the firesufficiently to allow them to continue searching the mine. On Saturday, January 21, they found thebodies of the missing miners.

Today, more than six months after these tragedies, a question still haunts many of the Sago miners’families:

Why was it possible to send mine rescue teams into the Alma Mine, within just a few hours after theemergency began, and past a burning fire — when the command center at Sago had refused to let rescueteams go in for nearly 10 hours after the explosion, at a mine where there was no fire, and then kept themon a tight leash instead of letting them head straight for the Two Left section?

Part of the answer is that there are some important differences between the two emergencies. There wasno explosion at Alma, and the mine liberated virtually no methane at all, so there was minimal risk of anexplosion occurring while rescuers were in the mine.49 The fire at the belt was not spreading out ofcontrol and therefore was not endangering either access or escape. Other than having to cope with smoke,

SAGO AND ALMA

79

continue reporting conditions from the minephone at 58 block, would it have been pos-sible to rush properly equipped rescuers into that point, taking the chance that thepossibility of fire and/or a second explosionwas low enough to justify the risk ?

It would have been asking a lot to expect thethree men to stay in the mine for many hours.Mine rescue teams were not on site in suffi-cient strength to launch an adequate rescueeffort until about eight hours after the explo-sion. And with no gas chromatograph on site,accurately assessing the fire risk would have

been difficult if not impossible until fairly latein the day on January 2.

As for the risk of a second methane ignition,whoever was calling the shots in the com-mand center would certainly have wanted toknow as much as possible about how muchmethane the mine liberated. This was not asdifficult as assessing the risk of fire. Fromeach quarterly inspection, MSHA had ahistory of the mine’s methane liberation. Sagoliberated very little methane — so little, in fact(approximately 100,000 cubic feet per 24hours47), that when MSHA District 3 Manager

heat, and carbon monoxide, for which they were trained and equipped, the rescue teams were not imper-iled by conditions in the mine. There was no reason to hold them back.

At Sago, in contrast, there was concern about what was not known. The gas readings obtained at theportal were subject to varying interpretations and the possibility that they might not reflect conditionsfurther inby. After about 9:30 a.m. on January 2, there was no one inside the mine to provide first-handreports on conditions. There was at least the possibility of a fire, and there was concern that the ventila-tion changes made in the first rescue attempt by mine superintendent Jeff Toler and others might havebeen pushing methane over such a fire, increasing the risk of a second explosion. Jeff Toler himselftestified to his anxiety about that possibility.50

So the two situations were not comparable. The rapid response at Alma does not necessarily mean thatthe more cautious response at Sago was fatally flawed.

But the Sago miners’ families are still right to raise these questions, because there are reasons to beconcerned about the differing responses.

What we know about human nature suggests that those in command at Alma (some of whom had alsobeen at Sago), as they confronted another emergency less than three weeks after the tragic outcome of theyear’s first disaster, had to have been aware of the need to organize the most aggressive response thatcould be justified under the circumstances — and they appear to have done that.

Their actions did not save the lives of the two missing miners. But they did demonstrate the value oftaking risks within reason. And that, in the end, is a large part of what mine rescue is all about.

80

Kevin Stricklin got the phone call telling himthat the Sago Mine had had an explosion, hethought the caller must have misidentifiedthe mine.48

Knowing that the mine was not a big meth-ane-producer, the command center mighthave figured that the initial explosion wouldhave burned out most of the methane andthat it would take a long time to build backup into the explosive range, even with themine’s ventilation impaired. So the risk ofsending mine rescuers in would have to beweighed against the advantage of sendingthem in as soon as possible — mainly toimprove the odds of finding men alive, ofcourse, but also to improve the odds of

completing the mission before a methanebuildup could imperil it.

It is possible — just possible — that such ascenario would have worked (even withouteyewitness reports on conditions at themouth of Two Left), and if everything hadgone just right, mine rescuers might —might — have reached the Two Left crew atsome point late in the afternoon on January 2,when some and perhaps all of the minerswere still conscious, as we can tell from theirnotes and from Randal McCloy’s recollection.But a mine rescue launched in such a waywould have put the rescue teams at high risk,no question about it. Although it seemscertain that they would have been willing to

REFUGES: A NECESSITY, NOT AN OPTION

Although escape should always be a miner’s first choice in a mine emergency, miners should not have tochoose between escape and barricading. They should have the option of safe refuge in a chamber capableof sustaining life for days if necessary.

Refuge chambers in coal mines are not a new or untried idea. In the 1970s, coal operators Frank andDusty Williams operated a mine in Upshur County, not far from where the Sago Mine is today, andvoluntarily built a rescue chamber in the mine. As the diagram shows, the chamber was excavated off the

intake air escapeway. It was 18 feet wide androughly 90 feet deep. Two concrete-filled 12-inchblock stoppings with offset doors provided blastprotection. The doors were far enough apart so thata stretcher could be passed between them. Thechamber was connected to the surface by two casedboreholes, one 6 inches in diameter and the other 8inches, which could be used to supply air, food, anda phone line. Dusty Williams remembers that thechamber was a break-even cost proposition “be-cause the coal paid for the work.”

One size does not fit all, in coal mining as else-where, so the refuge built by Upshur Coals should

UPSHUR COALS CORPORATIONEMERGENCY SHELTER HOLD

NO SCALE

6” CASED BOREHOLE 8” CASED BOREHOLE

CONCRETE FILLED 12” BLOCK

STOPPINGS

(INTAKE AIR ESCAPEWAY)

TO

1ST R

IGH

T M

AIN

S

TO

2N

D R

IGH

T M

AIN

S

81

take that kind of risk, it is unclear whetheranyone in the command center seriouslyconsidered such an option or would havebeen willing to shoulder the personal andprofessional responsibility for its failure.

It is not easy to consider this question, evenfrom a safe distance. Those who were in thecommand center on January 2 had to makelife-and-death decisions based on limitedinformation. Some might reject the idea thatan accelerated rescue operation could havehad any merit at all, recalling the rich historyof secondary explosions in coal mine disastersand believing that no responsible mineemergency manager should ever willinglysubject mine rescue teams to that kind of risk.

That view is understandable. But there isanother possible response, one that will hauntanyone closely involved with the Sagodisaster:

When a building is on fire and people aretrapped inside, a fire chief must sooner orlater make the gut-wrenching decision aboutwhether to send in his firefighters — knowingthat they have trained for the job and arewilling to accept the risks, and knowing fullwell that delays will almost certainly costlives. It is fair to ask why a coal mine shouldbe any different.

not be thought of as anybody’s ideal solution. It’s just a good example of miners’ ingenuity at work. Thereare other approaches. MineARC, an Australian manufacturer, custom-builds transportable steel refugechambers capable of accommodating up to 20 miners and supporting life for up to 36 hours. At a SolidEnergy underground coal mine in New Zealand, a transportable chamber is placed 300-600 feet from theface and is moved as coal is mined. Three other permanent chambers are installed in the mine, spacedabout 3,000 feet apart, connected by lifelines with directional cones. If miners are unable to escape themine, they can seek refuge in the chamber, communicate with the surface, replace their SCSRs if neces-sary, and then proceed to the next outby chamber using the lifelines. They can repeat the process, going tothe next “changeover station,” as the refuges are called, until they can safely exit the mine, or wait in oneof the chambers for instructions from mine rescuers.

There are numerous other refuge chamber designs, including one manufactured by Mine Safehouse,LLC, that was shown at the International Mining Safety and Health Symposium at Wheeling JesuitUniversity in April. The Safehouse structure is said to be capable of resisting fire at temperaturesexceeding 2000 degrees Fahrenheit and explosive forces up to 75 psi without conducting appreciable heatto the interior.

Obviously, careful consideration must be given to how and where a refuge chamber will be placed andhow to ensure that it will truly provide a safe shelter when needed. Exhaustive testing and designrefinements are a necessary part of getting refuge chambers into the underground mines of West Vir-ginia. But after Sago there can no longer be any excuse for postponing the day when every miner cancount on having a safe place to go to in an emergency if escape appears not to be an option.

82

SOURCES

1 Wilfong, 2/17/06 testimony, p. 59.

2 Stemple, 2/22/06 testimony, p. 14.

3 Stemple, 2/22/06 testimony, p. 13.

4 Urosek, 5/3/06 public hearing, p. 571.

5 Chewning, 3/28/06 testimony, pp. 17-18.

6 Martin, 3/23/06 testimony, p. 23.

7 Martin, 3/23/06 testimony, p. 24.

8 Hixson, 3/27/06 testimony, p. 25.

9 Klug, 3/24/06 testimony, p. 12.

10 Tucker, 3/28/06 testimony, p. 12.

11 Klug, 3/24/06 testimony, p. 24.

12 Lilly, 3/23/06 testimony, p. 26.

13 Jolly, 3/26/06 testimony, p. 12.

14 Hixson, 3/27/06 testimony, p. 14.

15 Lilly, 3/23/06 testimony, p. 23.

16 Lilly, 3/23/06 testimony, p. 13.

17 Thomas, 3/27/06 testimony, p. 10.

18 Hixson, 3/27/06 testimony, p. 14.

19 Jolly, 3/26/06 testimony, p. 14.

20 Jolly, 3/26/06 testimony, p. 14.

21 Lilly, 3/23/06 testimony, p. 27.

22 White, 3/28/06 testimony, p. 13.

23 Thomas, 3/27/06 testimony, p. 12.

24 Coleman, 2/21/06 testimony, p. 32.

25 Mine Safety and Health News, 4/3/06.

26 Mine Safety and Health News, 4/3/06.

27 Klug, 3/26/06 testimony, p. 27.

28 Martin, 3/23/06 testimony, p. 27.

29 Martin, 3/23/06 testimony, p. 28.

30 Klug, 3/26/06 testimony, p. 14.

31 Klug, 3/26/06 testimony, p. 14.

32 Bryant, 3/24/06 testimony, p. 20; federal andstate logs.

33 Hatfield, 5/2/06 public hearing, p. 224.

34 Hatfield, 5/2/06 public hearing, p. 224.

35 Bryant, 3/24/06 testimony, p. 21.

36 Bryant, 3/24/06 testimony, p. 22.

37 Bryant, 3/24/06 testimony, p. 22.

38 MSHA log 1/4/06.

39 Hatfield, 5/2/06 public hearing, p. 225.

40 Hatfield, 5/2/06 public hearing, p. 225.

41 Hatfield, 5/2/06 public hearing, p. 226.

42 Coleman, 2/21/06 testimony, p. 37.

43 Hatfield, 5/2/06 public hearing, p. 226.

44 Snavely, 5/2/06 public hearing, p. 229.

45 Martin, 3/23/06 testimony, p. 25.

46 Hixson, 3/27/06 testimony, p. 27.

47 Collins, 5/2/06 public hearing, pp. 160-161.

48 MSHA presentation to miners’ families,Buckhannon WV 3/9/06.

49 Joseph W. Pavlovich, memorandum to ThomasN. Bethell, 7/4/06.

50 Toler, 1/18/06 testimony, p. 15.

83

8 What happened to thethree shots?

‘I figured that they would bring that machine down and would have found us,would have drilled the hole in the right spot and would have took us out ofthere. That’s what I expected. I was expecting to hear shots fired... on top,above, and didn’t hear nothing. We banged and banged and banged, everyonedid.... but we never did know that I guess they... just didn’t have that machineto do that task.’

— Randal McCloy, Jr.1

Approximately three and a half years prior tothe Sago Mine disaster, at about 8:45 p.m. onJuly 24, 2002, miners working in the QuecreekMine, a small underground coal mine insouthwestern Pennsylvania, accidentallybroke through into an adjacent abandonedmine. Torrents of water rapidly flooded theQuecreek Mine, trapping nine miners. Threedays later they were all brought out alive,after rescuers drilled 240 feet down to withina few feet of the miners’ location and thenlifted them out of the mine, through theenlarged borehole, in a rescue capsule. Be-cause of nonstop television coverage, themiraculous rescue was seen by millions ofAmericans — including at least some of the12 miners who became trapped in the SagoMine on January 2, 2006.2

Tragically, the Quecreek rescue may helpexplain why there was such loss of life atSago. If the trapped Sago miners had knownthat no one was going to save them in theway the Quecreek miners were saved, wouldthey have abandoned their initial attempt tobreak out of the Two Left section on their

mantrip, after encountering debris from theexplosion blocking the tracks? They were insmoke, probably disoriented, and four of theirself-contained self-rescuers (SCSRs) wouldnot work for them. Randal McCloy, the solesurvivor of the crew, remembers their fore-man, Martin Toler, Jr., saying, “We’re going tohave to hit the section”3 — head back towardthe face, where they could hang curtain as abarricade and wait for the rescuers to listenfor them and then drill down to them.

If they had known that no one would belistening, would they have scrapped thatplan? We cannot know. What we do know,however, is that if they had been able toremove the debris from the ventilation over-cast which had fallen and was blocking thetracks, the way out of the Two Left sectionwould have been reasonably clear. It is truethat without enough functioning SCSRs to goaround, they might not have been able toremove the obstacles and then bull their wayout in their mantrip through the smoke andcarbon monoxide without some of themlosing consciousness along the way —

84

but they might have tried, if they had notthought, perhaps unconsciously, that aQuecreek rescue scenario was possible.

But why shouldn’t they have thought such arescue was coming? No one had told themnot to. The miners on Two Left did exactlywhat they had been told to do.

As every miner knows, miners are trained torespond to a mine fire, explosion, or otheremergency by first trying to get out of themine. If that fails, they are supposed tobarricade themselves against smoke and toxiccarbon monoxide and await rescue. As weknow from the testimony of Randal McCloy,the miners on Two Left went by the book.Led by their foreman, who was driving theirmantrip, they first tried to escape, and whenthey concluded that escape was not an option,they retreated to the safest place on thesection — the face, as far removed as possiblefrom the source of the smoke and fumes —to barricade themselves and await rescue.

The federal Mine Safety and Health Adminis-tration (MSHA) says it has distributed toevery miner in the United States a three-inch-square adhesive sticker, to be placed insidethe miner’s hard hat. The instructions printedon the sticker are very clear:

The trapped Sago miners tried to follow therules. Hearing nothing from the surface —no three shots — they used a sledgehammerto begin banging on a roof bolt, poundinghard as instructed, then resting for awhilebefore pounding again. It was exhaustingwork. They had to take their SCSRs out oftheir mouths in order to get enough air tokeep pounding4 — but that also meant theywere breathing carbon monoxide. Even so,for a long time they took turns pounding. Butthey never heard anything from the surface,let alone the five shots that were supposed totell them that help was on the way.

Meanwhile, on the surface, representatives ofInternational Coal Group were periodicallybriefing the miners’ families on the status ofthe rescue operation. West Virginia StateDelegate Bill Hamilton of Buckhannon,whose district includes the Sago Mine andwho was a friend of one of the missing miners— Junior Hamner — was present for some ofICG’s briefings. At the close of the publichearing on the disaster, he recalled that at oneof the briefings “the question was asked,‘Have you heard the miners pounding onanything?’ The answer, ‘No.’ The secondquestion, from the same miner’s family, was,‘Are you listening for sounds from the min-ers?’ Question answered, ‘Yes.’ But in testi-mony… we know that wasn’t true. Theyweren’t listening.”5

Nobody was listening. The explanation offeredafter the fact was that there was no need tolisten, because the command center had agood idea of where the miners were. That istrue up to a point, but only up to a point. Itwas likely that the trapped miners weresomewhere on the Two Left section. But thesection consists of multiple entries and is

WHEN ESCAPE IS CUT OFF

1. BARRICADE

2. LISTEN for

3 shots, then …

3. SIGNAL bypounding hard10 times

4. REST 15 minutes,then REPEAT signal until…

5. YOU HEAR 5 shots, whichmeans you are locatedand help is on the way.

85

more than half a mile in length from mouth toface. At the outset of the rescue effort it wasnot known that they were barricaded at theface. All that was known was that they werenot within shouting distance of where minesuperintendent Jeff Toler and other minershad hollered for them from the mouth of thesection before being forced to abandon thefirst rescue attempt in the early morninghours of January 2. As the command centerdiscovered many hours later — only afterlearning that the seals closing off the aban-doned area of the mine had been pulverized— there was also a possibility that the TwoLeft miners, while trying to escape, mighthave lost their way and ended up wanderinginto the no-longer-sealed section, possiblybarricading themselves a mile or more fromthe face of Two Left.

In other words, it would seem to have been agood idea to rush some seismic gear to Sago,set off the promised three shots, and listen forthe miners. But MSHA’s mobile SeismicLocation System never left its base nearPittsburgh. Why?

In print and on its website, MSHA describesits system in these words: “The seismiclocation system is truck-mounted and iscapable of detecting and locating the source ofseismic vibrations produced by trappedminers. Miners may generate seismic signalsby pounding on mine surfaces such as theroof, floor, or ribs. These signals are detectedby sensors installed either on the surface orunderground. The system is capable ofdetecting signals at a range of 1,500 feet, andcan monitor approximately one square mileover most mines (depending upon terrain).Helmet stickers describing how to signal havebeen distributed to all miners in the United

States. The system is highly mobile, and canbe airlifted.”6

This description is inaccurate. In reality, asMSHA personnel know, the seismic systemis old, outmoded, cumbersome and time-consuming to deploy, and it could not havebeen operational at Sago in time to be of anyhelp to the trapped miners.7

Ty Coleman, the man in charge of the rescueoperation for ICG, testified that he requestedMSHA’s seismic equipment at about 10:30a.m. on January 2, four hours after the explo-sion, and was assured by MSHA officials that“We’ve got the package coming to you,” as herecalled the conversation.8 If so, theymisspoke. The system stayed where it was.And the tragic fact is that there were goodreasons for leaving it there.

As MSHA officials know, setting up thesystem at Sago would have required exten-sive surveying above the mine, the construc-tion of a road capable both of accommodatingthe seismic truck and allowing it to movefreely from location to location above themine, the determination of precise coordi-nates in order to site the truck, and then thefiring of shots and — if no response washeard — the relocation of the truck to anothersite preparatory to firing off more shots.

The total elapsed time — from turning thekey in the ignition of MSHA’s parked 1970s-era truck until the first shots were fired —could easily have exceeded 15 hours.9 Even ifthe truck had left Pittsburgh at, say, 10:30 a.m.on January 2, when Coleman says he was toldit was on its way — and even if constructionof a road above the mine had started simulta-neously — the system probably would not

86

have been able to be operational until wellafter midnight, long after many if not most ofthem had lost consciousness because ofcarbon monoxide asphyxiation. By that timethey were far past the point of being able torespond to three shots, let alone listen for thefive that would tell them that the promisedhelp was on the way.

Much later — in fact not until more than amonth after the May 2-4 public hearing on theSago Mine disaster — MSHA acknowledgedthat, contrary to the upbeat information aboutits seismic system posted on the agency’swebsite, “MSHA’s experience has shown[that] application of the seismic technology isvery limited.”10 That is an understatement. Infact, the system has been deployed only twiceduring the past 20-plus years, and it did notsucceed in locating trapped miners on eitheroccasion.11 Although the system was trans-ported to the Quecreek mine rescue site, itwas not used even there, because othersurveying and drilling strategies proved to befaster and more accurate and efficient.

In any case, the miraculous Quecreek rescuewas made possible by a highly atypicalsituation. The Quecreek miners, althoughtrapped, were not breathing carbon monox-ide, because they had been trapped by aflood, not by an explosion or fire. They didnot run out of respirable air while they waitedfor three long days to be lifted out of themine. In the end they were saved by a goodjob of figuring out where they were and thenpinpointing the location of a borehole, butthat work took a long time. The Sago minersdid not have the luxury of time.

MSHA’s seismic system, developed nearlyfour decades ago, has been made obsolescent

both by its inherent limitations and by thedevelopment of superior technologies. AtSago, for example, ICG arranged for thedrilling of a borehole directly over the TwoLeft section using survey-grade GlobalPositioning System (GPS) instruments keyedto precise mine maps, with the entire processguided by cellphone by an engineeringconsultant who was in Atlanta at the time. Abulldozer built a road for a locally based drillwhile surveyors using wireless laptop com-puters in their pickup trucks determined thedrilling coordinates. Nobody had to drive anold rig from Pittsburgh to Upshur County.

Even so, the process of getting the boreholedrilled into the section was delayed by badweather — making it difficult for the survey-ors’ computers to pick up signals from mul-tiple earth-orbiting satellites, as GPS survey-ing requires — and the task needed nearly afull day to complete. That was record time,compared to the more than two days nor-mally required to survey and site a borehole,but it was still much more time than the TwoLeft miners had to spare. Moreover, difficultterrain prevented the surveyors from sitingthe drill so that it would penetrate close to theface of the section, where the trapped miners(if they had still been conscious and alert)might have heard or seen it.12 Instead, thedrill penetrated the mine with precision —punching into the middle of an entry ratherthan drilling into a solid block of coal aswould have happened if the coordinates hadbeen off by even a few feet — but it pen-etrated about 300 feet from where the minerswere barricaded.

During the May 2-4, 2006, public hearing onthe Sago Mine disaster, coal miner JohnHelms, who lost his brother Terry in the

87

disaster, questioned Kevin Stricklin, managerof MSHA’s District 3, about the cap stickersMSHA distributes to all miners. “I’ve workedin the coal mines for 37 years,” Mr. Helmssaid, “and got these little stickers where itsaid to barricade, beat on the roof, help willcome. I better throw them away — hadn’twe? We better come up with some otherplan.”

“I think we need to re-evaluate that, yes, sir,”Mr. Stricklin said.13

That was an honest answer. We do need to re-think the training given to miners in light ofthe time likely to be involved in a mine rescueoperation, even one that is conducted moreaggressively and at greater risk to the rescuersthan was the case at Sago. Perhaps even moreimportantly, we need to use the no-showfailure of MSHA’s seismic system to prod usinto some fresh thinking about better ways toprotect miners against a recurrence of whathappened to the waiting Two Left crew.

First and foremost in importance, of course, isprevention — starting with taking the neces-sary steps to systematically assess risks andminimize the chances that an explosion willoccur in the first place. If an area is to besealed, the supposedly “explosion-proof”materials used to seal it must in fact be able tocontain an explosion. It is equally urgent toensure that miners will not be failed by thesafety equipment that they carry or otherwisedepend upon. That means, among otherthings, equipping them with ample quantitiesof SCSRs that work, and equipping mineswith robust two-way communication systemscapable of surviving explosions and fires. Asdiscussed elsewhere in this report — and asrecognized by new state and federal legisla-

tion — there is no good reason to delay anylonger in getting such technology into themines of West Virginia and the rest of thenation.

That leaves the question of whether to investtime and money in a next-generation seismicsystem with greater capabilities than the onethat failed the Sago miners. It is a difficultquestion. We must always keep in mind thattime is the enemy when miners are trappedand barricaded, and the time likely to berequired to transport any MSHA-style cen-tralized seismic system from a remote loca-tion, even in-state, argues against relyingheavily on such a system. (In fact, based onthe evidence, our view is that the presentMSHA system should be mothballed. Inquir-ies made in the course of our investigationindicate that similar seismic technology hasnot been successfully deployed in SouthAfrica, Australia, or Canada, reinforcing ourbelief that MSHA’s system is unlikely ever toprove of value except, perhaps, in specificsituations not yet identified.)

On the other hand, researchers should beencouraged to develop a totally different kindof seismic system capable of much more rapiddeployment — such as a coiled-cable sensorsystem that could be laid out on the surface ofa mine to detect vibrations from someonepounding underground. Equipment could bekept in each MSHA district (or, depending onprice, required at every mine) and deployedquickly. Unless trapped miners can reach arefuge (discussed below and elsewhere in thisreport), pounding is all they can do — andthere seems little doubt that a commandcenter, hearing proof that trapped miners arealive, would be more likely to let rescue teamstake some risks to reach them.

88

Unless better seismic technology becomesavailable, an alternative is to take maximumadvantage of 21st-century mine surveyingtechniques. Given current GPS capabilities,for example, mine engineers know, as a mineadvances, exactly what the surface coordi-nates are for the working sections under-ground. In an emergency, surveyors needonly to be able to pick up geosynchronoussignals from multiple satellites to be able topinpoint the coordinates on the surface wherea borehole should be drilled.

That solves only part of the elapsed-timeproblem, however, since a road would stillhave to be built to the site and a pad con-structed to facilitate drilling. (Because of theneed to construct a pad, airlifting drills byhelicopter has limited applicability.) There-fore, even under the best circumstances,rescuing miners by borehole, as at Quecreek,should be thought of as a relatively remotepossibility, not a promise.

This reinforces the case for requiring emer-gency refuges in mines as the primary alter-native to escape — which should always bethe first choice. Suppose, for example, that arefuge chamber had been available on theTwo Left section — an explosion-hardenedshelter with an air supply, water, rations, andcommunication to the surface. Or suppose,even, that such a shelter had been located atsome distance, such as off of the main entries,equidistant between the One Left and TwoLeft sections.

With either of these scenarios, it is possible toimagine the miners of the Two Left crewemerging alive after even two or more days ofawaiting rescue. Even in the second scenario,knowing what we know of their failed at-

tempt to break out of the section on theirmantrip, it is at least possible to imagine thatthey would have struggled to remove thedebris blocking the track, knowing that theydid not have far to go to reach the safety ofthe refuge chamber.

Instead, they had no idea what they werefacing — and with no way of communicatingto the surface, they could not find out. For allthey knew, the explosion might have contami-nated the emergency escapeways all the wayto the portal, more than two miles from wherethey were. There was no way they would allhave made it that far on foot, especially withfour apparently nonfunctional SCSRs. So theirchoice was to risk losing some of the crew onthe way out or to stick together and make forthe section, as Junior Toler suggested. Theystayed together.

Randal McCloy honored the memory of hisforeman in these words: “He was, like,leading everything. He’s not the type of guywho’s pushy about anything, but he wasconcerned for everybody. He wanted to dothe right thing. He didn’t want to, like, get usstuck out in the woods with no rations,something like that.”14

Junior Toler did the right thing — but he andhis crew were failed by a mine rescue systemthat made promises it did not keep.

Above the mine, on the surface, a windingroad runs directly over the Two Left section,at about its midpoint. If you park a truckthere, get out, and walk cross-country in awesterly direction for a few hundred yards,you will be standing approximately above theface of the section, where Junior Toler and theTwo Left crew waited for some sign of rescue.

89

It is uneven terrain, and on January 2 it wouldhave been slick with rain. But a surveyor ormine engineer could have walked the terrain,accompanying someone equipped withexplosives. There are a lot of people in WestVirginia — miners, construction workers,highway engineers — who know how to useexplosives. If anyone in the command centerhad directed that three shots be set off aboveTwo Left to let the miners know that help wascoming, it could have been done.

But it was not done. And it is just not suffi-cient to say that the command center did notneed to set off shots because the probablelocation of the miners was known, or to argue

that setting off three shots would have servedno purpose since there was no system in placeto listen for the miners’ response. That missesthe point.

Someone in the command center must haveknown that the miners, if alive, would dowhat they had been trained to do — and whatall those cap stickers told them to do. But noone in the command center seems to havethought that it would be important to letthem know that help was on the way.

If the command center had a good idea wherethe miners were, five shots could have beenfired at the surface within, say, half an hour

Sago Mine, January 2, 2006

Two Left

One Left

Old Two Leftsealed area

EXPLOSIONORIGIN (approx)

Road

▲

Portal▲Omega Seals

90

after firing off the first three shots. The fiveshots would have told the miners they couldstop pounding, preserve energy, and possiblyprolong life by breathing as slowly and lightlyas possible. But no such message was sentto them.

Is this just a case of 20-20 hindsight, madepossible only by Randal McCloy’s survivaland his subsequent revelations about whathis crew did while awaiting rescue? No.

Those in the command center should havebeen thinking about all the what-ifs. They didnot. They failed in this regard, not becausethey are uncaring people — we do not believethat — but because they are part of an out-moded and inadequate mine emergencymanagement system that is in desperate need

of an overhaul. The system has grown stalewith disuse (and with the departure, partlythrough retirement, of some experiencedmine emergency managers). Its immediateneeds include greatly improved trainingrequirements and rigorous emergency-simulation drills covering, among otherthings, how to communicate with trappedminers until the day comes when they can bereliably contacted by explosion-resistant two-way wireless communication systems.

The Sago tragedy leaves no doubt that themine emergency management system mustchange, and fast. If another Sago disasterwere to occur tomorrow, would the outcomebe any different?

Right now, no one can be sure.

SOURCES

1Randal McCloy, Jr., 6/19/06 interview with federal and state investigators, transcript, pp. 63-65.

2Del. William Hamilton, closing statement, May 2-4, 2006, public hearing, transcript, p. 1237.

3Randal McCloy, Jr., 6/19/06 interview with federal and state investigators, transcript, p. 21.

4Randal McCloy, Jr., “To the families and loved ones of my co-workers, victims of the Sago Minedisaster,” 4/26/06 statement, p. 2.

5Del. William Hamilton, closing statement, May 2-4, 2006, public hearing, transcript, p. 1237.

6Jeffrey H. Kravitz and Robert G. Peluso, “The Mine Emergency Capabilities of the Mine Safety andHealth Administration,” undated, p. 5.

7 See, for example, the testimony of Kevin Stricklin and John Urosek at the May 2-4, 2006, publichearing, transcript, p. 599 (Stricklin) and p. 637 (Urosek).

91

8Harrison Tyrone Coleman, May 2-4, 2006, public hearing, transcript, p. 456.

9 J. B. Crawford, Letter to J. Davitt McAteer, re: Sago Mine Disaster Public Hearing, Additional MSHAInformation and Documents, June 22, 2006.

10 J. B. Crawford, Letter to J. Davitt McAteer, re: Sago Mine Disaster Public Hearing, Additional MSHAInformation and Documents, June 22, 2006.

11J. B. Crawford, Letter to J. Davitt McAteer, re: Sago Mine Disaster Public Hearing, Additional MSHAInformation and Documents, June 22, 2006.

12Gary Hartsog, 3/28/06 testimony, p. 35.

13May 2-4, 2006, public hearing, transcript, p. 602.

14Randal McCloy, Jr., 6/19/06 interview with federal and state investigators, transcript, p. 86.

92

1/4/06

Federal Mine Safety and Health Administra-tion (MSHA) announces its staff to investi-gate the Sago Mine Disaster.

1/9/06

MSHA announces it will conduct its investi-gation jointly with the West Virginia Office ofMiners’ Health, Safety and Training(WVOMHST)

1/9/06

Governor Joe Manchin III appoints J. DavittMcAteer as his special advisor on the SagoMine disaster and promises a public hearing.

1/11/06

McAteer briefs U.S. Senator Robert C. Byrdon the disaster.

1/12/06

Governor Manchin announces SenatorsCaruth, Kessler, Love and Delegates Caputo,Frederick and Hamilton will serve on thestate’s investigation team to work inconjunction with McAteer’s efforts.

1/15/06

Memorial service held for the Sago miners atWest Virginia Wesleyan College.

1/17/06

MSHA and WVOMHST begin voluntaryinterviews of witnesses.

1/18/06

MSHA appoints staff to conduct an internalreview of the agency’s response to the Sago

disaster and its pre-accident enforcement atthe mine.

1/19/06

Mine fire at Massey Energy’s Aracoma AlmaNo.1 mine kills two: Donald Bragg and ElleryHatfield.

1/20/06

U.S. Senators Rockefeller, Kennedy, Enzi, andIsakson meet with Sago families.

1/21/06

Investigators enter the Sago Mine to begincollecting physical evidence.

1/23/06

West Virginia Legislature passes mine safetybill SB-247.

1/23/06

U.S. Senate Appropriations Committee,Subcommittee on Labor, Health, and HumanServices, chaired by Senator Arlen Specterholds hearing on Sago disaster, with SenatorRobert C. Byrd leading the questioning.

1/24/06

MSHA seeks federal court order to grantUnited Mine Workers of America (UMWA)access to the investigation.

1/25/06

MSHA publishes in theFederal Register a Request for Information onemergency equipment and technology (e.g.,communication, robotics, refuge chambers.)

APPENDIX I: Synopsis of State of West Virginia

and federal activities following

the Sago Mine disaster

93

1/25/06

MSHA announces investigation team for fatalAracoma Alma fire.

1/26/06

Federal judge grants MSHA’s request for aninjunction against ICG to allow the UMWA toparticipate in the Sago investigation.

1/30/06

Esterhazy, Saskatchewan: 72 trapped Cana-dian potash miners survive 40 hours under-ground in a refuge chamber.

1/31/06

Richard Stickler, President Bush’s nominee forAssistant Secretary of Labor for Mine Safetyand Health, testifies at his Senate confirma-tion hearing.

2/1/06

Governor Manchin asks the state’s coal minesto temporarily cease operations until emer-gency safety reviews are conducted, follow-ing three separate mining accidents, in whichtwo miners were killed and another seriouslyinjured.

2/1/06

Senators Byrd and Rockefeller and Represen-tatives Rahall, Mollohan, and Capito intro-duce bills in their respective chambers ofCongress to enhance coal mine safety.

2/6/06

MSHA requests mine operators to “StandDown for Safety.”

2/7/06

Doug Conaway resigns as director ofWVOMHST.

2/8/06

Governor Manchin announces an Interna-tional Symposium on Mine Safety and Healthat Wheeling Jesuit University in April.

2/9/06

Governor Manchin appoints Jim Dean asacting director of WVOMHST.

2/10/06

WVOMHST asks the National Institute ofOccupational Safety & Health (NIOSH) toconduct explosion testing on Omega blocks.

2/11/06

MSHA reports it has received about 50proposals from manufacturers of under-ground communication and tracking devices;Consol Energy and Peabody Energy agree toassist in evaluating the devices.

2/13/06

Congressman George Miller (D-CA) holds aforum on mine safety, taking testimony fromfamily members of miners killed in the 2001Brookwood and 2006 Sago mine disasters.

2/15/06

MSHA announces plans to increase penalties;action is subject to formal rulemaking.

2/21/06

An explosion at a coal mine near San Juan deSabinas, Mexico traps 65 miners.

2/24/06

MSHA and WVOMHST conclude initialinterviews with witnesses.

2/27/06

WVOMHST files with the Secretary of Stateits emergency rule on SCSRs, communicationand other mine safety improvements.

94

2/28/06

Governor Manchin appoints a six-personlabor-management task force to assess avail-ability of communication systems and othertechnologies contained in new WVOMHSTrule.

3/1/06

House Committee on Education and theWorkforce, Subcommittee on WorkforceProtections, chaired by Charles Norwood (R-GA), conducts hearing on mine safety.

3/1/06

Governor Manchin announces that McAteerwill chair a public hearing on the Sago disas-ter (originally scheduled for mid-March) atWest Virginia Wesleyan College on May 2.

3/2/06

Senate Committee on Health, Labor andPensions, chaired by Senator Michael Enzi (R-WY), holds an oversight hearing on minesafety.

3/8/06

U.S. Senate Committee on Health, Labor andPensions approves Richard Sticker’s nomina-tion to be MSHA’s Assistant Secretary.

3/9/06

MSHA publishes an emergency temporaryrule in the Federal Register on mine evacuationprocedures and equipment; provisions aresimilar to the rules issued earlier byWVOMHST.

3/13/06

MSHA holds a public meeting to receiveinformation on communication and trackingtechnology, related to its January 25 requestfor information.

3/14/06

ICG meets with family members of thedeceased Sago miners and provides findingsof its on-site investigation, including thecompany’s conclusion that “the explosionwas ignited by lightning.”

3/15/06

Sago Mine resumes production.

3/23/06

MSHA and WVOMHST resume interviewsfor the Sago investigation, expanding theirinquiry to include the mine rescue effort.

3/27/06

NIOSH begins external and internal examina-tion of SCSRs recovered from the Sago mine.

3/30/06

Randal McCloy, Jr. is released from thehospital and continues his rehabilitation athome.

4/11/06

Family members receive transcripts of wit-ness interviews conducted by MSHA andWVOMHST.

4/15/06

NIOSH conducts the first in a series of explo-sion test of Omega block seals in its LakeLynn Experimental Mine.

4/18/06

NIOSH and MSHA host a workshop on mineescape planning and emergency shelters.

4/20/06

Governor Manchin convenes the first Interna-tional Symposium on Mining Health andSafety at Wheeling Jesuit University.

95

4/21/06

MSHA approves two handheld portableradios for use in underground mines, the firstsuch devices approved by the agency in 10years.

4/24/06

MSHA begins public hearings on its emer-gency temporary standard on mine evacua-tion procedures and equipment.

4/26/06

Randal McCloy, Jr. gives the Sago families aconfidential letter describing his and his co-workers’ experience while awaiting rescue.

5/2/06

Public hearing on the Sago mine disasterconvenes at West Virginia Wesleyan Collegein Buckhannon, WV; provides 3 days ofinquiry from family members, state andfederal officials and miners’ representatives.

5/5/06

West Virginia Board of Coal Mine Health andSafety files rule to establish two state-runmine rescue teams.

5/12/06

WVOMHST imposes a statewide moratoriumon the installation of Omega block seals andbegins a review of all alternative seals alreadyin place.

5/22/06

MSHA announces a moratorium on alterna-tive seals, a reassessment of the structuralintegrity of existing alternative seals, andrequirements to test the atmosphere behindseals.

5/29/06

West Virginia Mine Safety Technology TaskForce issues its report and recommendations.

6/8/06

MSHA provides an update on its testing andevaluation of communication and locatingdevices for underground miners.

6/9/06

WVOMHST files its emergency rule imple-menting recommendations of the Mine SafetyTechnology Task Force.

6/13/06

West Virginia adopts rule to establish twostate-run mine rescue teams; takes effect July13, 2006.

6/15/06

President Bush signs Mine Improvement andNew Emergency Response (MINER) Act intolaw.

6/15/06

NIOSH conducts the second in a series oftests of Omega block seals in its Lake LynnExperimental Mine.

6/19/06

MSHA, WVOMHST and McAteer interviewRandal McCloy, Jr.

6/29/06

Comment period ends on MSHA’s emer-gency temporary standard on mine evacua-tion procedures and equipment.

7/10/06

WVOMHST announces statewide SCSRinventorying and reporting system, effectiveAugust 15, 2006.

96

Acknowledgements

First, I wish to express my appreciate to our investigative team — Celeste Monforton,

Thomas N. Bethell, Joseph W. Pavlovich, Deborah Carpenter Roberts, Beth Spence,

and Dave Stuart — for their work in preparing this report. Special thanks also to

Patrick McGinley, Professor, West Virginia University School of Law; Suzanne Weise,

Esquire, and Jesse J. Wagner, Sr. whose detailed models of the Sago Mine have been

of immeasurable value in helping everyone involved with this investigation to visualize

the mine environment.

Special thanks to Wheeling Jesuit University and its President, Reverend Joseph R.

Hacala, S.J. Without his understanding, support and compassion, this report would not

have been possible. We offer sincere appreciation also to West Virginia Wesleyan

College for their professionalism and endless hospitality throughout our inquiry.

As I noted in the transmittal letter on the opening pages of this report, we owe a deep

debt of gratitude to the families of the victims of the Sago Mine disaster. They are a

constant and continuing source of inspiration and motivation to press for changes in

mine safety and the mining industry in the hope that other families will not have to

suffer as they are suffering. We also express our thanks to Randal McCloy, Jr., the sole

survivor of the Two Left crew, whose recollection of what he and his fellow miners

experienced has been of great value to our investigation.

We thank the students of Wheeling Jesuit University, West Virginia Wesleyan College,

West Virginia University, University of Charleston and Marshall University, who volun-

teered immediately following the disaster and in the months which followed. It makes

me proud as a West Virginian that these young people stepped up to help.

We want to express our gratitude to the many individuals — whether in state and

federal agencies, companies, professional or private life — whom we have called upon

for advice, guidance and expertise in the preparation of this report. And we are grate-

ful to the hundreds of individuals who have contacted us with ideas and suggestions

about how to improve mine safety and health in West Virginia and across the nation.

This investigation has been supported by the State of West Virginia and its Governor,

Joe Manchin, III. We thank Governor Manchin and his staff for their support and

assistance during this investigation and preparation of the report.

The list of names that follows encompasses, we hope, most of those who have been

involved in one way or another with our work over the past six months. Inevitably,

however, there are bound to be omissions. We apologize for that, and hope that

anyone whose name should have appeared here will be brought to our attention

for future correction.

J. Davitt McAteer

97

Aimee Adams • Lynda Anderson and family • Jonathan Andrew • Kent Armstrong • Pamela

Jubin Balch • Simon Barber • Richard D. Begley • Judy Bennett, Russell Bennett and family

• Jason S. Bentley • Jamie Berg • Jodi Besenyei • Jacek Bielawa • Phil Bounds •

Wojciech Bradecki • LeeAnn F. Brown • Melissa A. Brown • Tony Bumbico • U.S. Senator

Robert C. Byrd • Brent Bodine • Premier Lorne Calvert • Pam Campbell • U.S.

Congresswoman Shelley Moore Capito • State Delegate Mike Caputo • State Senator Don

Caruth • Mary Ellen Cassidy • Elizabeth Chamberlin • Skye M. Chernicky • Gene Cilento •

Kevin Cimino • Peggy and Aaron Cohen • Robert Cohen, MD • Doug Conaway • Kimberly

A. Cottrell • Tammy Crites • Sean Daniel • Taylor A. Daugherty • James Dean • Christo

de Klerk • Mary Doane • Jerry Donahue • Earl Dotter • Carol and Doug Duffield • Robert

Edge • Chuck Edwards • Miranda Elkins • Susan Elliott • Yvonne Farley • Alan Fine •

Mike Foletti • Brian France • State Delegate Eustace Frederick • Bruce G. Freedman, MD •

Jimmy Joe Gianto • Richard T. Gillespie • Bobby Godsel • Michael Griffin • Cara J. Group •

Wanda Groves and family • Father Joseph R. Hacala, SJ • State Delegate Bill Hamilton •

Professor Vivian Hamilton • Debbie Hamner and family • Larry Harrah • Bennett K. Hatfield

• William R. Hayden • Monte Hieb • Amber Helms and the Helms family • Jerry Henderson

• F.W. Hermann • Justin Hershberger • David Hinkins • Brenda Hissam • Jessica L. Holley

• Tad Hudkins • Craig Hyre • Justin Jack • Darwin F. Johnson • Yvonne Johnson •

Shenna L. Johnston • Lula Bell Jones and family • Jarrod Jordan • State Senator Jeff

Kessler • State House Speaker Bob Kiss • Sam Kitts • Jill Kriesky • Kimberly N. Leezer •

Samantha Lewis • Pete Lilly • Alisa Lively • Tricia Lollini • Kathleen M. Long • Kathryn G.

Lough • State Senator Shirley Love • Bernard Lown, MD • Joseph P. McAteer • Timothy O.

McAteer • Randal and Anna McCloy • Professor Joyce McConnell • Pastor Ed McDaniels •

Professor Marjorie McDiarmid • Professor Pat McGinley • Darrell V. McGraw, Jr. • Andrew

Meek • Ann and Dan Merideth and family • David Michaels • Mine Safety and Health

Administration • U.S. Congressman Alan B. Mollohan • Virginia Moore • Keith Moran •

Thomas E. Moran • Christina Morgan • Zachery M. Morton • Paul Myles • Kraig R. Naasz •

National Institute for Occupational Safety and Health • National Mining Association • Blaire

Nuzum • James O’Dell • Thomas J. O’Neill • Carolyn Ostand, RN • Brenda Pinnell •

Jennifer Powell • Andrea Pyles • U.S. Congressman Nick J. Rahall, II • Bill Raney • Paul

and Cindy Rank • Katy Ratai • James Rau • Lawrence H. Roberts, MD • U.S. Senator John

“Jay” Rockefeller • Joseph R. Romano • Shelly and Mike Rose • Chad Rosepappa • Neil

Roth • Robert O Rupp • Allison E. Sharpe • Elizabeth H. Short • Bob Skinner • Malcolm

Smith • Mayor Nicholas Sparachane • Catherine E. Squires • Duwane Squires • Ben Stout

• David Stuart • Catherine E. Swigert • Tony Szwilski • Ryan A. Thorn • Mary Lou Toler

and family • Senate President Earl Ray Tomblin • U.S. Department of Labor • United Mine

Workers of America • Kyla R. Vala • Jesse J. Wagner, Sr. • Charlotte Weaver and family •

Charlotte Weber • Suzanne Weise • Dusty Williams • Frank Williams, Jr. • Kimberly N.

Williams • Ann Wilson • Pam Winans and family • Tom Wingler • WV Office of Miners

Health Safety and Training • WV Board of Coal Mine Health & Safety • WV Coal Association •

WV Department of Homeland Security • Wheeling Jesuit University Board of Directors • Evan

M. Wolfe • Kevin G. Younkin • Gary Zamel • Paul Ziemkiewicz