producing natural gas from coal - schlumberger/media/files/resources/oilfield_review/ors03/... ·...

8 Oilfield Review

Producing Natural Gas from Coal

John AndersonMike SimpsonNexen Canada LtdCalgary, Alberta, Canada

Paul BasinskiEl Paso ProductionHouston, Texas, USA

Andrew BeatonAlberta Geological SurveyEdmonton, Alberta

Charles BoyerPittsburgh, Pennsylvania, USA

Daren BulatSatyaki RayDon ReinheimerGreg SchlachterCalgary, Alberta

Leif ColsonTom OlsenDenver, Colorado, USA

Zachariah JohnPerth, Western Australia, Australia

Riaz KhanHouston, Texas

Nick LowClamart, France

Barry RyanBritish Columbia Ministry of Energy and MinesVictoria, British Columbia, Canada

David SchoderbekBurlington ResourcesCalgary, Alberta

For help in preparation of this article, thanks to ValerieBiran and Tommy Miller, Abingdon, England; Ian Bryant,Leo Burdylo, Mo Cordes and Martin Isaacs, Sugar Land,Texas, USA; Matthew Chadwick, Worland, Wyoming, USA;Ned Clayton, Sacramento, California, USA; AndrewCarnegie, Abu Dhabi, United Arab Emirates; Steve Holditch,College Station, Texas; Lance Fielder, Cambridge, England;Stephen Lambert and Mike Zuber, Pittsburgh, Pennsylvania,USA; Harjinder Rai, New Delhi, India; John Seidle, SprouleAssociates Inc., Denver, Colorado, USA; and Dick Zinno,Houston, Texas. Thanks to Willem Langenberg, AlbertaGeological Survey, and Ken Childress, photographer, for providing the outcrop and rig photographs, respectively (above).AIT (Array Induction Imager Tool), APS (AcceleratorPorosity Sonde), CemNET, ClearFRAC, CoilFRAC, DSI(Dipole Shear Sonic Imager), ECLIPSE Office, ECS(Elemental Capture Spectroscopy), ELANPlus, FMI (Fullbore

Formation MicroImager), LiteCRETE, MDT (ModularFormation Dynamics Tester), OFA (Optical Fluid Analyzer),Platform Express, RST (Reservoir Saturation Tool), SFL(Spherically Focused Resistivity), SpectroLith and StimMAPare marks of Schlumberger.

Natural gas in coal formations is an important resource that is helping address the world’s growing energy

needs. In many areas, market conditions and technological advances have made the exploitation of this

resource a viable option. The unique characteristics of coalbed reservoirs demand novel approaches in well

construction, formation evaluation, completion and stimulation fluids, modeling and reservoir development.

1. For more on the history of coal exploitation (July 1, 2003):http://www.bydesign.com/fossilfuels/links/html/coal/coal_history.html and http://www.pitwork.net/history1.htm

2. Leach WH Jr: “New Technology for CBM Production,”Opportunities in Coalbed Methane: A Supplement to Oiland Gas Investor, December 2002, Oil and GasInvestor/Hart Publications, Houston, Texas, USA.Schwochow SD: “CBM: Coming to a Basin Near You,”Opportunities in Coalbed Methane: A Supplement to Oiland Gas Investor, December 2002, Oil and GasInvestor/Hart Publications, Houston, Texas, USA.

Autumn 2003 9

With global oil production moving from plateauto decline, worldwide reserves of natural gas takeon added importance. Increasingly, gas is viewedas a vital alternative energy source because it isplentiful and burns cleaner than other fossilfuels (see “A Dynamic Global Gas Market,”page 4). In mature, high-demand markets, theindustry is looking at nonconventional gassources, such as shale gas, low-permeabilitysandstones and coalbed methane. These uncon-ventional gas accumulations cannot be exploitedin the same way as conventional reservoirs, presenting challenges to both operators and service companies.

Natural gas from coal seams is an importantpart of the world’s natural gas resource.Improved methods of evaluating coals are nowavailable from new logging measurements and sampling devices. Lighter cements, with the effective use of additives, minimize damageto sensitive coalbed methane reservoirs.Nondamaging fracture-stimulation fluids andinnovative hydraulic fracturing designs are beingused to improve gas and water flow to the well-bore. Optimized artificial-lift techniques areachieved through the use of intelligent softwareto promote quick and efficient dewatering ofcoals. Advanced technologies and industry expe-rience applied worldwide are having a positiveimpact on coalbed methane (CBM) development.

This article discusses CBM reservoirs, alsoknown as coalbed natural gas (CBNG) or coalseam methane (CSM) reservoirs. First, we reviewthe history of coal exploitation. Next, we discussthe geologic processes that led to the formationof coal, how coals generate and store natural gas,and what makes CBM reservoirs so different fromtraditional clastic and carbonate gas reservoirs.Finally, case studies from around the worlddemonstrate the industry’s use of various tech-nologies to evaluate and develop CBM reservoirs.

Minds, Mines and WellboresHumans have appreciated the energy value ofcoal for thousands of years. Early use of coal infires, dating back to 200 BC, has been confirmedin ancient Chinese records. There is even evidence that Stone-Age inhabitants in Britaincollected coal; archeologists have found flintaxes implanted in coal seams. The earliest coalfinds exploited by humans were used to supple-ment firewood supplies and were likely found atthe surface, along rock outcrops near streambanks. The first evidence that humans dug forcoal was found in regions where firewood wasscarce. Mining techniques evolved from theprimitive method of finding an exposed coalseam along hillsides, then digging into the hill as

far as possible to extract the coal. When the oper-ation became too dangerous, these early coal dig-gers would move to another location along theoutcrop.1 From excavation sites in Britain, it hasbeen determined that as early as 50 AD, Romansmined coal to fuel heating systems and smeltingoperations. Eventually, pits were dug to accessthe coal.

Modernization of mining methods, includingroom and pillar, and longwall mining techniques,enabled larger and deeper operations, exposingmine workers to a variety of hazards. One signifi-cant hazard in coal mining is methane gas—a by-product of the coal thermal maturation process that becomes a serious problem in deepermines. Mine operators alleviated dangerous conditions in the subsurface by using mine-ventilation techniques. Air pumped into a minethrough mineshafts and ventilation pipes provided oxygen to workers and dissipated thepoisonous and explosive methane. Mining compa-nies also drill coal-degasification wells into coalseams to liberate methane gas prior to mining thecoal. Modern ventilation and degasification techniques paved the way for a safer and moreproductive mining industry. Coal mining in manyareas is still not completely safe, so degassing the mines using wellbores ahead of mining operations is an extremely important techniqueto help reduce the number of mining accidents.

Coal became the energy behind the industrialrevolution in Western Europe and across theworld, and remains an important resource today.However, there is more to the value of coal than just burning it for heat and electricity; thenatural gas that once was merely a hazard can be produced and distributed like conventionalnatural gas, providing a clean-burning fuel.

Drilling for Coalbed Natural GasCoal degasification in mines was first attemptedin England during the 1800s, and it is reportedthat the coal gas was used for lighting the streetsof London. The first CBM well to develop gas as aresource was drilled in 1931 in West Virginia,USA. For more than 50 years, CBM drilling activ-ity remained low. In 1978, the US governmentpassed the Natural Gas Policy Act. This legisla-tion allowed companies to receive higher pricesfor natural gas produced from tight gas reservoirs,gas shales and coal seams. In 1984, the US gov-ernment offered tax credits for developing andproducing unconventional reservoirs. Originallyset to expire in 1990, the tax credits wereextended two more years because of their positiveimpact on drilling activity. After the tax creditsexpired in 1992, low gas prices caused concernsabout the economics of CBM development.

Gas price is not the only factor affecting theviability of CBM production. Accessibility to gas-transportation infrastructure and technical issuesrelated to CBM production, for example low initialgas-production rates, high water-production ratesand disposal issues, must also be considered. Thepositive impact of accessibility to adequatepipeline capacity can be seen in portions of theRocky Mountains, USA, where the expansion ofthe Kern River Pipeline in May 2003 has signifi-cantly improved gas-production economics.

Today, CBM development is having an impacton the North American gas market. Annual production from 11 coal basins in the US nowexceeds 1.5 Tcf [42.9 billion m3], or 10% of theannual US gas production (above).2 ProvenCBM reserves—17.5 Tcf [501 billion m3]—nowmake up 9.5% of US total gas reserves, and thetotal US CBM in place is estimated at 749 Tcf

1000 10,000

8000

6000

4000

2000

0

12,000

14,000

16,000

18,000

20,000

0

2000

1200

1400

1600

1800

800

600

400

200

1981Year

1983 1985 1987 1989 1991 1993 1995 1997 1999 2001

Annu

al C

BM p

rodu

ctio

n, B

cf

Prod

ucin

g CB

M w

ells

> US coalbed methane (CBM) production (blue) and number of producing wells (red).

[21.4 trillion m3]. About 100 Tcf [2.9 trillion m3]are thought to be recoverable (below).3

Increased gas prices, the continued expansion ofthe natural gas transportation system and recentadvances in oilfield technologies have helpedmake CBM wells more profitable. Through theyears, operators and service companies havegained valuable knowledge from miningresearch, and practical experience from drillingactivity induced by the US tax credits.

As operators drilled and produced CBM reser-voirs, it became clear that coal reservoirs behavedifferently from basin to basin, and even withinbasins. This behavior largely guides the applica-tion of different technologies within a basin orfield. In many CBM areas, operators have reducedtotal exploitation costs while increasing gas recov-ery by prudent application of new technology.

Canada has just started to produce gas fromCBM reservoirs and estimates its in-placereserves to be 1287 Tcf [36.8 trillion m3].Australia started producing CBM in 1998 andplaces its total reserves at 300 to 500 Tcf [8.6 to14.3 trillion m3]. Worldwide, the total CBM in-place reserves are estimated to be between3500 and 9500 Tcf [100 and 272 trillion m3].4 By2001, 35 of the 69 coal-bearing countries had

investigated CBM development but, just as inNorth America, the pace of future developmentwill depend on economics (next page, top).

From Peat to CoalThe formation of coal starts with the deposition oforganic material from plants, creating peat. Peatis formed by continued subaqueous deposition ofplant-derived organic material in environmentswhere the interstitial waters are oxygen-poor.Distinct environments allow the accumulation,burial and preservation of peat, including swampsand overbank areas that may or may not bemarine influenced (next page, bottom). In thegeologic past, most peat is thought to have formedin deltaic or marginal marine environments.

Coalification, or the conversion of peat intocoal, occurs at different rates in different envi-ronments. Biochemical degradation initiates thecoalification process, but with burial, increasingoverburden pressures and subsurface tempera-tures cause physicochemical processes that continue coalification. As water, carbon dioxideand methane are released, the coal increases in rank, which is a measure of maturity. Coals are divided into rank stages and include, in order of increasing rank: sub-bituminous,

high-volatile bituminous, medium-volatile bitu-minous, low-volatile bituminous, semi-anthracite,and anthracite coals. Although coals containsome inorganic minerals, they are composedlargely of macerals, or vegetal compounds, ranging from woody plants to resins.

The three general categories of macerals arevitrinite, liptinite and inertinite. Vitrinite refersto woody plant material, like trunks, roots,branches and stems. Liptinite macerals corre-spond to the more resistant parts of the plant,such as spores, pollen, waxes and resins.Inertinite macerals represent altered plantmaterial and are less structured. These maceralshave a greater carbon content from oxidationprocesses that occurred during deposition, forexample the burning of wood or peat in fires.Maceral data reflect the basic makeup of coalsand therefore help geologists determine CBMreservoir potential.

An Unconventional ReservoirFrom the time of deposition, coal is differentfrom other kinds of reservoir rock. It is composedof altered vegetative material—macerals—thatfunction as both hydrocarbon source and reservoir. It is inherently fractured from the

10 Oilfield Review

Tertiary

Tertiary-Cretaceous

Cretaceous

Jurassic

Triassic

Pennsylvanian and Permian

Mississippian

0

0 200 400

100 200 300

600 800 km

400 500 miles

NorthernAppalachian

Richmond

CentralAppalachian

WesternWashingtonCoal Region

North CentralCoal Region

Cahaba/Coose

BlackWarrior

Gulf Coast

Illinois

Forest City

Bighorn

GreaterGreen River

Wind River

Uinta

KaiparowitsPlateau

San Juan

PowderRiver

HannaCarbon

Denver

Arkoma

Cherokee

Piceance

Raton

> US basins containing coalbed methane reserves. Major coal basins are shown with the associated periods of coal deposition.

Autumn 2003 11

coalification process, which forms vertical fractures, or cleats. Coal cleats are classified geo-metrically with the primary, more continuouscleats called face cleats and the secondary, lesscontinuous cleats called butt cleats.

Genetic classification of coal fractures is alsocommon. Endogenetic fractures, or classic cleats,are created under tension as the coal matrixshrinks due to dewatering and devolatilization dur-ing coal maturation. These cleat sets are orthogo-nal and nearly always perpendicular to bedding. Incontrast, exogenetic fractures form due to tecton-ism and therefore regional stress fields dictatetheir orientation. Shear fractures, oriented 45º tothe bedding planes, also are observed.

In virtually all coalbed reservoirs, cleats arethe primary permeability mechanism. Like con-ventional reservoirs, coals can also be naturallyfractured. In deeper coal seams, higher over-burden stresses can crush the coal structure andclose the cleats. In such locations, subsequentnatural fracturing tends to be the main perme-ability driver. Understanding the cleating and

3. Nuccio V: “Coal-Bed Methane: Potential and Concerns,”U.S. Geological Survey, USGS Fact Sheet FS–123–00,October 2000. http://pubs.usgs.gov/fs/fs123-00/fs123-00.pdf

4. Olsen TN, Brenize G and Frenzel T: “ImprovementProcesses for Coalbed Natural Gas Completion andStimulation,” paper SPE 84122, presented at the SPEAnnual Technical Conference and Exhibition, Denver,Colorado, USA, October 5–8, 2003.

>Worldwide coalbed methane activity. By 2001, 35 (red dots) of the 69 coal-bearing countries had investigated CBM development.

> Peat-forming environments. Peat is formed by continual subaqueous deposition of organic matter inenvironments where waters are poorly oxygenated. The accumulation, burial and preservation of peatoccur in a range of environments that include swamps and overbank areas. These may or may not bemarine influenced. (These photographs of the Loxahatchee River, Florida, USA, are from the SouthFlorida Water Management District Web site: www.sfwmd.gov/org/oee/vcd/photos/hires/hilist.html)

natural-fracture systems in coals is critical during all facets of CBM reservoir development.

Methane generation is a function of maceraltype and the thermal maturation process. As temperature and pressure increase, the rankof the coal changes along with its ability to generate and store methane (left).5 Also, eachmaceral type stores, or adsorbs, different vol-umes of methane. In addition, coal can storemore gas as its rank increases.

Conventional sandstone and carbonate reser-voirs store compressed gas in porosity systems.Methane is stored in coal by adsorption, a pro-cess by which the individual gas molecules arebound by weak electrical forces to the solidorganic molecules that make up the coal. Toassess how CBM wells might produce over time,the sorptive capacity of crushed coal samples aretested and desorption isotherms are constructed(below). Desorption isotherms describe the rela-tionship between pressure and adsorbed gas content in the coal at static temperature andmoisture conditions. Coal’s ability to storemethane largely reduces the need for conven-tional reservoir-trapping mechanisms, making itsgas content—which is related to coal rank—andthe degree of cleating or natural fracturing theoverriding considerations when assessing anarea for CBM production potential.

12 Oilfield Review

Biogenic methaneNitrogen

Carbon dioxide

Lignite Sub-bituminous Bituminous Anthracite Graphite

Thermally-derivedmethane Volatile matter

driven off

Incr

easi

ng g

as v

olum

e

Gas Generation as a Function of Coal Rank

Increasing coal rank

> Gas generation in coal. As temperature and pressure increase, coal rankchanges along with its ability to generate and store methane. Through time,dewatering and devolatization occur, causing shrinkage of the coal matrixand creation of endogenetic cleats.

Coal versus Sandstone–Gas Content versus Pressure

Coal

Sandstones

Gas

cont

ent,

scf/

ton

(coa

l equ

ival

ent)

200

100

0

300

400

500

600

5000 1000 1500 2000 2500 3500 45003000 4000Pore pressure, psia

4% porosity to gas6% porosity to gas8% porosity to gasCoal isotherm

1000

Abso

rbed

gas

con

tent

, scf

/ton

(dry

, ash

-free

)

1200

800

600

400

200

00 200 400 600 800 1000

Pressure, psia

AnthraciteMedium-volatile bituminousHigh-volatile bituminous AHigh-volatile bituminous B

Methane Sorptive Capacity versus Coal Rank

> Sorptive capacity of coal. As coal maturity increases from bituminous to anthracite, the sorptive capacity of coal increases. Tests conducted on coalsamples to relate adsorbed gas to pressure—under isothermal conditions—assess how CBM wells might produce over time. The plot showstypical responses in bituminous and anthracite coals (left). The gas storage capacity of coal can be significantly greater than that of sandstones (right).

Production time

Prod

ucin

g ra

te, M

scf/

D or

STB

/D

Stage lStage ll Stage lll

Gas

Water

Well “dewatered”

> Coalbed production characteristics. During Stage I, production is domi-nated by water. Gas production increases during Stage II, as water in thecoal is produced and the relative permeability to gas increases. DuringStage III, both water and gas production decline.

Autumn 2003 13

This storing ability gives coals unique early-time production behavior that is related to des-orption, not pressure depletion. Coals maycontain water or gas, or both, in the cleat and natural fracture systems, and gas sorbed onto theinternal surface of the coal matrix. Any water present in the cleat system must be produced toreduce the reservoir pressure in the cleat systembefore significant volumes of gas can be pro-duced. Dewatering increases the permeability togas within the cleats and fractures, and causesthe gas in the matrix to desorb, diffuse throughthe matrix and move into the cleat system, result-ing in CBM production profiles that are quiteunique (previous page, middle).

Initial production is dominated by water. Asthe water moves out of the cleats and fractures,gas saturation and production increase and waterproduction falls. When permeability to gas even-tually stabilizes, the coal is considered dewateredand gas production peaks. From this point, bothwater and gas production slowly decline, with gasbeing the dominant produced fluid. The speed atwhich the reservoir dewaters depends on severalfactors, including original gas and water satura-tions, cleat porosity, relative and absolute perme-ability of the coal, and well spacing.

Some CBM wells produce dry gas from thestart. For example, some wells in Alberta andBritish Columbia, Canada, and the under-pressured portion of the San Juan basin are com-parable to conventional reservoirs and producewater-free at irreducible water saturation. Drygas coalbed production typically declines fromthe start, exhibiting Stage III behavior.

As with all gas reservoirs, the permeabilitycontrols production and largely dictates theamount of gas reserves in coal seams. Local vari-ations in cleat and natural-fracture conductivityand density—how closely cleats or fractures arespaced—lead to wide variations in well perfor-mance within some areas of development (aboveright). For example, 23 wells in a field in theBlack Warrior basin, USA, with similar coal thick-nesses and original gas contents, were drilled andcompleted identically, at equal well spacings, butshow diversity in production performancebecause of the local variations in cleat conductiv-ity—permeability. Also, in this basin, cleat andnatural-fracture conductivity are greatly affectedby the stress on the reservoir. Field-test data con-firm the inverse relationship between closurestress and coal permeability; increasing closurestress from 1000 to 5000 psi [6.9 to 34.4 MPa]decreased permeability from 10 to 1 mD.

The unconventional properties and produc-tion performance of coalbed reservoirs, includinghigh initial water production and low initial gasproduction, are largely responsible for the rela-tively slow uptake in CBM reservoir developmentaround the world. However, the collective knowledge and experiences of the industry inexploiting this resource are showing results inincreased CBM production.

Investigating a New Resource in IndiaAfter reviewing the major coal-bearing basins inIndia, the Oil and Natural Gas Corporation(ONGC) concluded that the Jharia basin, 250 km[155 miles] northwest of Calcutta, had the bestpotential for coalbed natural gas production.Three pilot wells were drilled through thePermian-age Barakar formation, which containsup to 18 clearly identifiable coal beds, each from1 to 20 m [3 to 66 ft] thick. The second pilot holewas cored and logged with high vertical resolu-tion lithodensity, neutron and resistivity mea-surements from the Platform Express integratedwireline logging tool, FMI Fullbore FormationMicroImager, DSI Dipole Shear Sonic Imager andECS Elemental Capture Spectroscopy tools.Fullbore cores were obtained in many of thecoals and were sent for proximate analysis, rankdetermination and adsorbed gas content.6 The

logs were analyzed for these same parametersand for cleat porosity.

The first step was proximate analysis from thelithodensity, neutron and gamma ray logs. Theselog measurements have widely differentresponses to the various coal components andcan resolve them well. The main uncertainty liesin the response parameters of ash, since it maycontain varying amounts of quartz, clay, calcite,pyrite and other minerals.7 The parameters ofvolatile matter—mainly organics, wax, carbondioxide [CO2] and sulfur dioxide [SO2]—andfixed carbon are reasonably similar for the bitu-minous and anthracite coals of interest. In theJharia well, results of the log analysis were in

5. Zuber M and Boyer C: “Evaluation of Coalbed MethaneReservoirs,” prepared for the University of Oviedo, Spain.Holditch-Reservoir Technologies Consulting Services,Pittsburgh, Pennsylvania, USA, May 24–25, 2001.

6. Proximate analysis is the term used for the identificationof the major fractions of the coal, taken as moisture,volatiles, fixed carbon and ash. These fractions haveusually been determined by progressively heating andthen burning crushed samples and observing the volumeof the different fractions removed at each stage untilnothing is left but ash. Proximate analysis is distinct fromultimate analysis, in which the weight percent of differ-ent elements is determined.

7. Ash is the inorganic constituent, derived from mineralmatter, that remains after proximate analysis.

Cum

ulat

ive

gas

prod

uctio

n, M

scf

175,000

150,000

125,000

100,000

75,000

50,000

25,000

00 10 20 30 40 50 60 70 80

Time, months

Production from 23 CBM Wells

> Local well-performance variations in a group of 23 similar wells in a field in the Black Warrior basin,USA. In this area, the differences are attributed to local changes in cleat and natural-fracture perme-abilities. The plot shows cumulative gas production through time for each of the 23 wells.

good agreement with core data (above). The ECSdata added detailed information on the composi-tion of the ash and improved the estimate of totalash in washed-out coals, where the density andother logs were more affected by the borehole

(see “The Elements of Coal Analysis,” page 16).The next step was to estimate the volume of

adsorbed gas in each seam. Ideally, this would bederived directly from logs. However, the effect ofadsorbed gas on the response parameters of coal

is small and there are not enough independentmeasurements to solve reliably for gas.Traditional coal-industry techniques determinegas content from cores, and in their absence, byestimating the rank of the coal from proximateanalysis and the gas content from rank, pressure,temperature and a suitable adsorption isotherm.The American Society of Testing and Materials(ASTM) ranks coals by the percentage of volatilematerial after normalizing to dry, ash-free coal.Slightly different ranking criteria were used inJharia and were applied to both core and log data.

With logs providing information on intervalswhere core data were missing, ONGC was able tostudy the quality of the different coal seams. Theaverage coal rank increased with depth, but witha probable change in trend half-way down thesection (next page, top). The change in trend ismost likely related to a major fault seen on theFMI data at this depth. Coal rank and proximateanalysis can also be entered into a suitable sorp-tive capacity transform to determine the gas inplace within each coal seam.8

Cleat porosity was estimated by four differentmethods: from the porosity seen by microresistiv-ity measurements, by the separation of deep andshallow laterolog curves, by the quantity and typeof mineralization seen by the ECS tool, and fromthe shear-wave anisotropy measured by DSI data.When the borehole is in gauge, the microresistiv-ity measurement gives the most accurate results,and is used to calibrate the ECS and DSI data. Inwashed-out coals, the ECS log is least affected byhole rugosity, while the DSI and microresistivitylogs can be affected more severely. The estimateof cleat porosity adds information on flow capac-ity to that already obtained on gas volume. Thesedata helped ONGC decide which seams to test,whether to develop this resource and how best toaccomplish this.

Huge Reserves and Progress in CanadaCanada’s estimated 1287 Tcf of probable in-placeCBM reserves lie primarily in the provinces ofBritish Columbia and Alberta, and can be dividedinto three main areas, the Alberta foothills, theAlberta plains and the British Columbia foothills.Coals from these areas vary in rank, gas contentand accessibility. Canadian coal experts main-tain that coal permeability is the main driver ofCBM reservoir potential. For this reason, much ofthe focus when assessing CBM reservoirs inCanada is on understanding cleats and naturalfractures, both in outcrop and in wellbores.

Alberta contains vast amounts of coal dis-tributed throughout the southern plains, foothillsand mountains. Originally deposited in relatively

14 Oilfield Review

API

GammaRay

Log Coal Rank

Dept

h, m

Core Coal Rank

ELAN volumes

% Volatile matterVolatiles + fixed carbon

X060

X070

Sem

i-ant

hrac

ite

Low

-vol

ume

bitu

min

ous

Med

ium

-vol

ume

bitu

min

ous

High

-vol

ume

bitu

min

ous

Sub-

bitu

min

ous

17 24 36 40

10 500 1vol/vol0 API

Gamma Ray

6 16in.Hole Size

Ash

Fixed Carbon

Volatile Matter

Clay

Quartz

Water

MoistureCaliper > Bit Size

150

High Resolution

Density Neutron

g/cm3

2.75

1.35

0.90

1.00

0.05

0.45

1.00

1.00

vol/vol

Photoelectric EffectCapture Cross Section

barn/cm3

12

0.2

0.5

0.5

400

20

0

0

Ash

Fixed Carbon

Volatile Matter

Moisture

> An example of proximate analysis and coal rank determination from logs in India. In Track 1, thecaliper indicates that the hole is moderately washed out but still smooth. Track 2 shows good agreementbetween the log-derived proximate analysis, using the parameters given in the table, and core-derivedanalyses. Track 3 compares coal rank from logs, after applying a vertical average, with coal rank fromcore. Coal rank is determined by the proportion of volatile material in the dry, ash-free coal, using thecutoffs shown (bottom).

Autumn 2003 15

flat-lying peat swamps, organic matter wasburied by sediments derived from the west andgradually coalified with increasing heat and pres-sure after burial. Coals were subsequently folded,faulted, uplifted and partially eroded, resultingin the present distribution of coal across theplains. Coal-bearing strata gently dip westwardtowards the mountains, where the coals arefolded and abruptly turn towards the surface tobe reexposed in the foothills.

Coal seams occur within distinctive horizonsof the upper Cretaceous Scollard, HorseshoeCanyon and Belly River formations, and withinthe lower Cretaceous Mannville group strata inthe Alberta plains. Coal is also found within thePaleocene Coalspur formation and the MistMountain formation of the Jurassic-CretaceousLuscar/Kootenay groups in the Alberta foothills(below left). Individual coal seams vary in thick-ness from less than 1 meter [3 ft] to more than6 meters [20 ft]. Groups of coal seams are sepa-rated by 10 to 50 m [30 to 160 ft] of rock. Mostcoals at shallow depths—less than 1000 m[3300 ft]—in the plains are sub-bituminous tohigh-volatile bituminous rank. Coals in theAlberta foothills generally are more mature, withranks from high-volatile to low-volatile bitumi-nous. Alberta plains coals have more predictablecleat characteristics than foothills coals inAlberta and British Columbia because of theirlimited deformation.

Permeability, formation pressure and reser-voir fluid saturation are critical in identifyingareas suitable for CBM development. Commonmethods used to measure permeability in coals,such as injection and falloff testing, often yieldinconsistent results because the cleat permeabil-ity can be a function of injection pressure. Testintervals may be disturbed by drilling fluids andcan be damaged by cementing, breakdown andstimulation fluids, causing adverse effects on test results. Ambiguities occur for a variety ofreasons, including inflation of coal cleats andfractures, two-phase permeability and wellbore-storage effects.

8. The theory of Langmuir relates the gas volume adsorbedon ash-free coal to pressure at a given temperature andto two factors that depend on temperature and coalrank. Various researchers have correlated these factorswith the results of proximate analysis, so that theadsorbed gas volume can be estimated from logs. SeeHawkins JM, Schraufnagel RA and Olszewski AJ:“Estimating Coalbed Gas Content and Sorption IsothermUsing Well Log Data,” paper SPE 24905, presented at theSPE Annual Technical Conference and Exhibition,Washington, DC, USA, October 4–7, 1992.

Vola

tile

mat

ter,

wt %

(dry

, ash

-free

)

25

20

15

10

30

40

45

50

35

Increasing log depth, m100 m

X50

X55

Dept

h, m

deg0 90

BeddingTrue Dip

0 120 240 360

FMI Static ImageRes. Cond.

Orientation North

Fault

Fault location

> Percent of volatile material—dry and ash free—and coal rank versus depth for the Jharia well. Log-derived data (red curves) and core-derived data (blue dots) are shown only in the coal seams. Thecore-derived data, in particular, suggest a change in trend (blue line) probably associated with a faultobserved on the FMI image (inset) and in other data at that depth.

(continued on page 20)

0

0 200 400 km

100 200 300 miles

Mannville Group

Horseshoe CanyonGroup

Scollard Formation

Belly River Group

Kootenay Group

Luscar Group

Coal Zones withCBM Potential

Calgary

Edmonton

Low- and medium-volatile bituminous coalsHigh-volatilebituminous coals

Sub-bituminous coal

Lignite

Distribution of Coal by Rank

Edmonton

Calgary

Alberta

> Alberta coals. Maps show the distribution of major coal seams (left) and coal rank (right) in Alberta.

16 Oilfield Review

In the simplest technique of proximate analy-sis from logs, the bulk density is interpretedfor ash content, which is then correlated withthe other proximates for each rank of coal.Addition of the neutron, gamma ray and photo-electric logs makes the analysis more generaland less dependent on local correlations.Unfortunately, some coals tend to wash outwhile drilling, leading to oversize boreholesand large borehole effects on the logs. In addi-tion, the composition of the components, inparticular ash, can vary, creating some uncer-tainty in the parameters to be used in interpretation.

An alternative technique is based on elemen-tal analysis from neutron-induced gamma rayspectroscopy. Both the ECS Elemental CaptureSpectroscopy sonde and the RST ReservoirSaturation Tool device estimate the quantity ofminerals in the coal. The advantage of neu-tron-induced gamma ray spectroscopy is thatthe majority of the signals of interest arisefrom elements in the formation and are there-fore unaffected by the borehole. In addition,the components of the ash can be more pre-cisely defined from the mineralogy.

Neutron-induced gamma ray spectroscopytools emit high-energy neutrons that are thenslowed down and captured by elements in the

borehole and the formation. During capture, agamma ray is emitted with an energy that ischaracteristic of the element. A detector mea-sures the gamma ray spectrum, or the numberof gamma rays received at the detector at eachenergy level. This energy may be degraded byscattering in the formation, but there is suffi-cient character in the final spectrum to recog-nize the peaks caused by different elements.The first processing step is to calculate theproportion, or relative yield, of gamma raysfrom each element by comparing the measuredspectrum with the theoretical spectrum ofeach individual element (next page). A mathe-matical inversion provides the percentage ofthe principal contributors, such as silicon, calcium, iron, sulfur and hydrogen.

The yields are only relative measuresbecause the total signal depends on the envi-ronment, which may vary throughout thelogged interval. To obtain the absolute ele-mental concentrations, additional informationis needed. The principle of oxide closurestates that a dry rock consists of a set ofoxides, the sum of whose concentrations mustbe unity.1 Measuring the relative yield of allthe oxides allows the calculation of the totalyield and the factor needed to convert thetotal to unity. This normalization factor will

then convert each relative yield to a dryweight elemental concentration.

Finally, the SpectroLith lithology processingtechnique transforms elemental concentra-tions into mineral concentrations using a set ofcorrelations based on the study of more than400 core samples from different clastic envi-ronments.2 The results are expressed as thedry weight percentage of clay, coal, accessoryminerals such as pyrite and siderite, and theaggregate of quartz, feldspars and micas. Whilethere may be local variations in these correla-tions, the major advantage of this technique isthat it is automatic, with no user intervention.This contrasts with standard methods for claydetermination that depend heavily on user-selected parameters.

Coals are easily identified by their highhydrogen concentration. Quantifying theamount of fixed carbon, volatile material andmoisture in coal is more difficult and requirestwo assumptions. First, there are other sourcesof hydrogen that must be considered, includingwater in the cleats, clay water and moisture inthe formation, and in the borehole, unless thewell was drilled with air. Since these form aconsistent background, they can be subtractedto give the hydrogen concentration in coal.Second, different types of coal have different

The Elements of Coal Analysis

San Juan Basin, USA

y = -0.834x + 75.471R2 = 0.997

y = -0.171x + 24.034R2 = 0.944

0 10 20 30 40 50 60Ash, wet weight %

70

y = 0.005x + 0.495

Prox

imat

e co

mpo

nent

, wet

wei

ght %

MoistureVolatilesFixed carbon

40

30

20

10

0

50

60

70

80India

MoistureVolatilesFixed carbon

Prox

imat

e co

mpo

nent

, wet

wei

ght %

y = -0.7762x + 75.575R2 = 0.8662

y = -0.2245x + 24.286R2 = 0.3507

Ash, wet weight %

y = -0.0014x + 0.1816

40

30

20

10

0

50

60

70

80

0 10 20 30 40 50 60

> Proximate analysis based on ash content. Excellent correlations have been found with data from three wells in the Fruitland coal interval in theSan Juan basin (left). The correlations from the Jharia well in India are satisfactory for fixed carbon but poor for volatile matter (right).

Autumn 2003 17

hydrogen contents. However, in a given area orformation, this can be sufficiently consistent toallow a conversion from hydrogen concentra-tion to coal percentage.

Data from the ECS tool allow a quick and automatic proximate analysis at the wellsite. The total ash content is simply obtained fromits components, namely quartz, clay, carbonatesand pyrite, while the amount of fixed carbonand volatile material can be estimated fromcorrelations with ash content (previous page).

1. In practice, the process is not so straightforward.First, we measure elements, not oxides, but nature ishelpful since the most abundant elements exist in onlyone common oxide, for example quartz [SiO2] for sili-con [Si]. Thus for most elements there is an exactassociation factor that converts the concentration ofthe element to the concentration of the oxide. Second,although the ECS tool measures a majority of the mostcommon elements, there are exceptions, the most

important being those of potassium and aluminum.Luckily, the concentration of these elements isstrongly correlated to that of iron, so that they can beincluded in the oxide association factor for iron.

2. Herron S and Herron M: “Quantitative Lithology: AnApplication for Open and Cased Hole Spectroscopy,”Transactions of the SPWLA 37th Annual Symposium,New Orleans, Louisiana, USA, June 16–19, 1996, paper E.

Inelastic

H

Si

Cl

Fe

Gd

Induced Gamma Ray Spectra

X00

X50

X00

X50

HClGdTiSFeCaSi

Inversion (Spectral Stripping)

Coal

GdTiFeSCaSi

Oxide Closure

Pyrite (wt %)

Carbonate (wt %)

Sand (wt %)

Coal (wt %)

Clay (wt %)

SpectroLith Model

m

Energy

Coun

ts

> The interpretation steps for obtaining mineralogy from gamma rays. Thedetector receives a spectrum of gamma rays that is compared with standardsfor each element to obtain their relative yields. The yields are converted intoelemental concentrations by applying a normalization factor computed fromthe oxide-closure model. Finally, the SpectroLith model estimates mineral percentages from elements.

18 Oilfield Review

Many such correlations have already beenestablished from core data for specific areasor formations.3 Alternatively, the ECS miner-alogy can be combined with other log data inan ELANPlus computation. The resultingproximate analysis is enhanced by thedetailed ash description from the ECS sonde,and by the ability of lithodensity and neutrondata to distinguish between fixed carbon andvolatile matter.

The more detailed ECS mineralogy alsohelps identify the degree of cleating. The

presence of calcite and pyrite indicates a well-developed cleat system in which the flow ofwater has caused secondary mineralization.However, large quantities of calcite and pyritesuggest that the cleats have been filled or thatthe coal is of low grade. Quartz and clay havealso been observed in cleats, but large vol-umes of these minerals and a large total ashvolume indicate a lower-ranked coal. Suchcoals will have lost less water and volatilematter during coalification and will thereforehave fewer cleats.4 These observations can be

used to identify well-cleated coal by, for exam-ple, calcite percentages between 2 and 7%, andpyrite percentages between 0.5 and 5%. Poorlycleated coals have total ash percentages above45%, clay percentages above 25% and quartzpercentages above 10%. Mineral percentagesthat fall between those of well-cleated coalsand poorly cleated coals indicate partly cleatedcoals.5 The rules and cutoffs can vary by areaand should be established locally from produc-tion data.

Core Mineral Ash

0 0.4wt %

0 0.4wt %

Mineral Ash

Mineral Ash

0 1wt %

Core Fixed Carbon

0 1wt %

Core Fixed Carbon0 1wt %

Core Moisture

Volatiles

Fixed Carbon

Mineral Ash

Moisture

0 500

Core Desorbed Gas Content

0 50

Max. Bed Gas

0 50

Min. Bed Gas

MMcf/acre

MMcf/acre

0 500scf/ton

Max. Gas Content

0 500scf/ton

Min. Gas Content

0 500scf/ton

Gas Content (Hawkins Study)

Gas Content Range

Bed Gas Content Range

0.2 2000ohm-m

SFL Resistivity

0 0.1ft3/ft3

Openhole Resistivity

0 0.1ft3/ft3

RST Cleat Porosity

Well Cleated

Partly Cleated

Poorly Cleated

scf/ton

> A typical coal evaluation using neutron-induced gamma ray spectroscopy from the RST ReservoirSaturation Tool. Tracks 1 and 2 show proximate analyses from logs and core. Track 3 shows gas content and cumulative gas content from core and from logs using two different transforms. One is the Langmuir Rank equation developed by Hawkins et al, reference 8, main text. The other is a localequation based on ash content, temperature and pressure. Track 4 indicates the cleat intensity.

Autumn 2003 19

3. Hawkins et al, reference 8, main text.4. Law BE: “The Relationship Between Coal Rank and

Spacing; The Implications for the Prediction ofPermeability in Coal,” Proceedings of the InternationalCoalbed Methane Symposium, Vol. 2, Birmingham,Alabama, USA, (May 17–21, 1993): 435–442.

5. Ahmed U, Johnston D and Colson L: “An Advancedand Integrated Approach to Coal FormationEvaluation,” paper SPE 22736, presented at the 66thSPE Annual Technical Conference and Exhibition,Dallas, Texas, USA, October 6–9, 1991.

R2 = 0.9025

2.0

1.8

1.6

1.4

1.81.2

2.0 2.2 2.4 2.6 2.8Carbon/oxygen ratio

Open

hole

bul

k de

nsity

Dept

h, ft

X450

X500

6 16 1 3g/cm3

0 10

Density Derived from Carbon/Oxygen Ratio

Photoelectric Factor

Openhole Formation Density

1 3g/cm3

Coal Indication from Openhole Density

Caliper

in.

0 200API

Gamma Ray

> Comparison of density from an openhole log (red) and that derived from the RST carbon/oxygen ratio(black), after making the best-fit correlation shown in the plot (inset). The openhole density suggests acoal at X447, but the carbon/oxygen data show this to be incorrect and to be caused instead by thewashout seen on the caliper.

The coal rank and gas content can be esti-mated based on proximate analysis. Cleatintensity indicates permeability and hence productivity. Thus, neutron-induced gammaray spectroscopy, in combination with otherlogs, provides a continuous record of themajor factors needed to evaluate a coal seamand any surrounding sands shortly after thewell has been drilled (previous page).

Elemental analysis has an additional role incased holes, where the RST carbon/oxygenratio is the most accurate logging method for

identifying coals. This technique is particularlyuseful in wells drilled for deeper targets thathave been cased over the coal-bearing zoneswithout recording an openhole density log. Thecarbon/oxygen ratio is calibrated to coal den-sity, using data from other wells in the area(above). The other elemental yields can beinterpreted as already described after allowingfor the effects of casing and cement on the silicon and calcium concentrations.

Nexen Canada Ltd. has run successful tests onshallow plains coal seams using the MDT ModularFormation Dynamics Tester device (next page).After pumping out drilling fluid, the MDT packermodule can withdraw reservoir fluid from isolatedcoal seams at near-virgin conditions. The tool provides accurate flow rate and pressure informa-tion, and measures the properties of the recoveredfluids in real time. Pressure-transient analysis canbe applied to the pressure response to determinecoal permeability. Bottomhole shut-in pressurereduces the problem of wellbore storage that canmask the formation response in pressure-transient analysis. Nexen Canada has found that

the MDT device is cost-effective and minimizesuncertainties inherent in other coal-permeabilitytesting methods.

Some of the Mannville coals of the Albertaplains are thin-bedded, as seen in a BurlingtonResources Canada well FMI image (above). Here,the bulk density log seems to respond to heavyminerals like pyrite in the coal matrix. These areseen as conductive specks on FMI images, givinganomalously high density peaks that cause somepotential errors in net-coal estimates. The higherresolution of the FMI tool allows more reliablenet-coal thickness measurement.9

Abundant folds and thrust faults related toLaramide deformation characterize the complexstructural geology of the British Columbia andAlberta foothills. The present minimum horizontalstress direction runs northwest-southeastthroughout much of the foothills area, roughlyparallel to the outcrops, although local stressvariations are indicated in borehole-breakoutstudies. In the Alberta plains, recent studies by

20 Oilfield Review

API0 150

Gamma Raymm125 375

Bit Sizemm125 375

Caliper

0 120 240 360 0 120 240 360

deg0 90

1000 2650

BeddingTrue Dip

kg/m3

X49.6

X49.8

X50.0

X50.2

X50.4

X50.6

X50.8

X51.0

X51.2

FMI Static ImageResistive Conductive

FMI Dynamic ImageResistive Conductive

Dept

h, m

Orientation North Orientation North

Density

Coal unresolved by thedensity log

5-cm coal layer

Probable subvertical cleats

Pyrite inclusions affectthe density measurement

> High-resolution measurements in thin-bedded coals. Many coals are thin-bedded and may not beidentified with standard measurements. The FMI Fullbore Formation MicroImager tool has a verticalresolution of 0.2 in. [0.5 cm], which allows analysts to image thin coals. Track 1 contains gamma ray,caliper and borehole orientation data. A comparison between the density log and the FMI static imageis displayed in Track 2. The FMI tool clearly identifies the thin coal at X50.0 m, where the density logdoes not. Pyrite inclusions that dramatically affect the density at X51.0 m appear as dark spots on theFMI image. Track 3 contains the FMI dynamic image, and Track 4 displays dip information.

9. Schoderbek D and Ray S: “Applications of FormationMicroImage Interpretation to Canadian CoalbedMethane Exploration,” presented at the CSPG–CSEGAnnual Convention, Calgary, Alberta, Canada, June 2–6, 2003.

(continued on page 23)

Autumn 2003 21

15157

15017

14877

14737

14597

14457

14317

14177

14037

13897

13757

13617

13477

13337

13197

13057

12917

12777

12637

12497

12357

12217

12077

11937

11797

11657

Time, s

Events

500

13.00

0.00

0

Pressure/Temperature/Resistivity

Pressure (kPa) - PASG

Temperature (°C) - PAQT

Resistivity (ohm-m) -

Pump Out Volume (C3) - POPV

GasDetection

FluidFraction OFA Fluid Density

Events

High

Medium

Low

Water

Oil

Test Position#2

1500

15.00

0.00

80,000Mud

Fluid Color

Bit Size

mm125 375

mm125 375

API

Caliper

0 150

Gamma Ray

0.2 2000ohm-m

AIT Resistivity 90-in.0.2 2000ohm-m

0.2 2000ohm-m

AIT Resistivity 30-in.


0.2 2000mD1000 1500

0

0.75

0.6

500

10

0.15

1000kPa

Spherical PermeabilityHydrostatic Pressure Formation Pressure

True

ver

tical

dep

th, m

kPa

Photoelectric Effect

Neutron Porosity

0

XX25

XX15Hydrostaticgradient

Testposition #1

Testposition #2

Density Porosity

m3/m3

m3/m3

925

915

905

895

885

875

865

855

845

835

825

Pres

sure

, kPa

0 0.05 0.10 0.15 0.20 0.25Spherical time function

Test Position #2Specialized Analysis Plot-Spherical Flow Buildup

k sph = 1.289 mDp int = 916.9 kPa

Delta

P &

der

ivat

ive

grou

ps, k

Pa

1e+03

1e+02

1e+01

1e+00

1e-011e-01 1e+00 1e+01 1e+02 1e+03

Delta T, s

PressureDerivative

Test Position #2Flow Regime Identification Plot-CBM Buildup

Bit Size

mm125 375

mm125 375

API

Caliper

0 150

Gamma Ray

0.2 2000ohm-m

AIT Resistivity 90-in.0.2 2000ohm-m

0.2 2000ohm-m



0.2 2000mD1000 1500

0

0.75

0.6

500

10

0.15

1000kPa

Spherical PermeabilityHydrostatic Pressure Formation Pressure

True

ver

tical

dep

th, m

kPa

Photoelectric Effect

Neutron Porosity

0

XX25

XX15Hydrostaticgradient

Testposition #1

Testposition #2

Density Porosity

m3/m3

m3/m3

> Pressure and permeability from the MDT Modular Formation Dynamics Tester device. Nexen Canada Ltd. ran the MDT tool ona well to test coal seams in the Alberta plains. The MDT test position 2 can be located on the log (top). Hydrostatic pressures areplotted in Track 1, along with the gamma ray and caliper data. Results from the spherical flow buildup permeability analysis (middleleft) are plotted in Track 2, along with the resistivity data. Buildup data were also used to identify a spherical flow regime (bottomleft). Formation pressure determined from the buildup analysis is plotted in Track 3, along with porosity and lithology information.The OFA Optical Fluid Analyzer plot (right) shows pressure, temperature and pumpout volume during sampling, and fluid recov-ery changes during the test. Drilling mud was recovered initially (brown), then water (blue) with some possible small shows ofgas (white).

22 Oilfield Review

0 20deg

API0 150

Gamma Ray

125 375mm

Caliper 2

125 375mm

Caliper 1

Bit Size

125 375mm

0 10


1 0m3/m3

Neutron Porosity

1000 2650kg/m3

Density

2e-05 cm

Fracture Aperture

0.2 2000ohm-m

0.2 2000ohm-m

deg0 90

UnconformableBed Boundary

True Dip

deg0 90

Drilling-Induced FractureTrue Dip

deg0 90

Resistive FractureTrue Dip

deg0 90

Partial Open FractureTrue Dip

deg0 90

Cross BeddingTrue Dip

deg0 90

Bed BoundaryTrue Dip

After-Frac Survey

120 240 3600

Background

CoalScandium

Antimony

Iridium

Background

CoalScandium

Antimony

Iridium

XX70

Mea

sure

d de

pth,

m

XX75

XX80

XX85

Gamma Ray45 to 75 API

Gamma Ray<45 API

Borehole Drift

Coal 0.2



Fault drag

Fractures in coal Perfs

Perfs

FMI Dynamic Image

Resistive Conductive

Orientation North

> Analysis of Alberta plains coal seams. A fault was identified during the FMI image interpretation of this Burlington well at a depth of XX79.5 m(Track 4). Faults and associated fracturing have a direct impact on the permeability of coal seams. Gamma ray and caliper data are displayed inTrack 1 with borehole orientation. Track 2 contains porosity and lithology information. Fracture apertures exceeding 0.01 cm [0.004 in.] were calcu-lated from FMI data and are displayed with resistivity data in Track 3. Track 4 contains the dynamic FMI image from which bedding and fracturesplanes were picked. Track 5 shows the dip plots from the interpretation of Track 4. An after-frac survey is included on the right to demonstrate thevertical growth of hydraulic fractures from the perforated coals. The presence of radioactive tracers below the perforations indicates downwardfracture growth.

Autumn 2003 23

the Alberta Energy and Utilities Board (AEUB)using regional geological data, and drilling andcompletion records indicate stress variationbetween upper Cretaceous-Tertiary and lowerCretaceous rock sequences.10 In addition, imagedata from the FMI tool have shown faults in theseareas (previous page). After-fracture surveys wererun to evaluate how hydraulic fractures propagatethrough the coals and surrounding rock.

In the foothills of northeast British Columbia,the Cretaceous Gates and Gething formationscontain the thickest coal resources. Coals inthese formations are exposed in the Peace Rivercoal field, along northwest-trending outcrops,where they are mined. In the southeast corner ofBritish Columbia, coal is contained in theJurassic-Cretaceous Mist Mountain formation,which outcrops in the front range of the RockyMountains in the Elk Valley, Crowsnest andFlathead coal fields.

The Gething formation contains over 20 m[65 ft] of cumulative coal in the Pine River area.The formation thins regionally southeastwardbut maintains cumulative coal thicknesses ofabout 6 m. A 1980 report of coal exploration inthe northern part of the Gething trend providesinformation on gas contents from drill holes. Thedata indicate high gas contents—up to19.5 m3/tonne [620 scf/ton] at a depth of 459 m[1506 ft] in at least one hole. Gething coal rankgenerally decreases to the east and spans thebituminous range.11 Face cleats within theGething coals in the north trend northwest-southeast and, under the current stress regime,may be closed.12

The Gates formation thins to the northwestand its coal reserves are not as widespread asthose in the Gething formation. Where coal ispresent in the Gates formation, it normally contains four coal seams with an average totalthickness of 15 to 20 m [49 to 66 ft]. In 1996,Phillips Petroleum drilled four wells to test theGates formation coals at a depth of 1300 to1500 m [4270 to 4920 ft]. Gas contents measuredin these wells were promising and ranged from6.3 to 29.2 m3/tonne [202 to 935 scf/ton],although measured permeability was low. Facecleats in Gates coal trend northeast-southwestand may be perpendicular to the present minimum stress direction. It is therefore reason-able to surmise that the face cleats in the Gatesformation may be open.

The outcrop exposures in the Peace Rivercoal field have given geologists insights into the interrelationships between deforma-tion, cleat development and present stress fields, and their relationship to coal permeability. The

combination of depth and deformation may have significantly reduced the permeability in coalseams in the Gething and Gates formations.Intraseam shearing of these coals is thought tohave diminished coal permeability.

Coal outcrops provide extensive informationon stresses and coal-fracture systems. In the sub-surface, many operators rely on borehole imagingto determine the degree of cleating and naturalfracturing within the coals; in some wells, shearfractures can be observed using borehole images(above). Burlington Resources Canada and their

10. Bell JS and Bachu S: “In Situ Stress Magnitude andOrientation Estimates for Cretaceous Coal BearingStrata Beneath the Plains Area of Central and SouthernAlberta,” Bulletin of Canadian Petroleum Geology 51, no.1 (2003): 1–28.

11. Marchioni D and Kalkreuth WD: “Vitrinite Reflectanceand Thermal Maturity in Cretaceous Strata of the PeaceRiver Arch Region, West-Central Alberta and AdjacentBritish Columbia,” Geological Survey of Canada, OpenFile Report 2576, 1992.

12. Bachu S: “In Situ Stress Regime in the Coal-BearingStrata of the Northeastern Plains Area of BritishColumbia,” Sigma H. Consultants Ltd. Invarmere BC,Report for the Ministry of Energy and Mines, BritishColumbia, 2002.

British Columbia Foothills Coal

API0 150

GammaRay

mm125 375

Bit Size

Caliper

mm125 375

0 90

0 120 240 360

deg

Bedding True Dip

Face Cleat

deg0 90


Orientation North

Mea

sure

d de

pth,

m

XX20

XX21

XX22

Face cleatButt cleat

Shear fractureBedding

Foothills Coal

Alberta Plains Coal

API0 150

GammaRay

mm 375Bit Size

Caliper

mm 375

deg0 90

Bedding True Dip

Face Cleat

0 120 240 360 deg0 90

Mea

sure

d de

pth,

m

Orientation North


XX59

XX60

125

125

Plains Coal

Shearfractures

Face cleat

Butt cleat

Face cleat

> Comparison of FMI images from the Alberta plains coal and British Columbia foothills coal. Theimage of a plains coal shows clear face- and butt-cleat development (top left). The images of thefoothills coal help geologists identify significant shear fracturing (top right). Outcrop exposures ofAlberta plains and British Columbia foothills coals show bedding planes, face and butt cleats, andshear fractures. Features are marked on the outcrop photographs. The foothills coal (bottom right)shows extensive shear fractures, while the plains coal does not (bottom left). Shear fracturingdegrades coal permeability.

partners have acquired FMI data to determinecleat and fracture directions, as well as present-day stress orientation. This information is used inwell planning and aids in the evaluation ofhydraulic fracture stimulation behavior andeffectiveness (left). Drilling-induced fracturesand borehole breakouts indicate orientation ofin-situ stresses. High-quality borehole images ofnatural fractures facilitate interpretation ofpaleostress orientations and fracture apertures.Deviated wells are drilled perpendicular to thedominant fracture set using FMI informationfrom nearby wellbores or uphole log data.Borehole images also are used to orient anddepth-match cored intervals, particularly inzones of poor core recovery (below left).

In addition to borehole imaging, shear andcompressional acoustic-velocity data have longbeen used with other petrophysical measure-ments like bulk density, porosity and shale vol-ume to derive rock elastic properties and todetermine closure-stress profiles for input tohydraulic fracture designs.13 Although thesemethods have been routinely used in westernCanada for several years, their application incoals is a recent phenomenon.

Multiarray induction logs can provide invasion profiles and qualitative comparisons of cleating in coals. In Canada, geologists and petrophysicists with Burlington andSchlumberger are investigating a method toassess coal permeability by examining drilling-fluid invasion using AIT Array Induction ImagerTool data. The AIT device provides resistivitymeasurements at five depths of investigation,ranging from 10 inches to 90 inches, and withvertical resolutions of 1, 2 and 4 feet. The inva-sion profile is calculated using a model with afully flushed zone of diameter Di, followed by azone of transition to the uninvaded formation atdiameter Do. The model has been used to com-pute the invasion profile in two contrasting wells,a low-permeability foothills CBM test well and ahigher permeability plains CBM well. Both wellswere drilled with fresh mud systems, providing agood resistivity contrast between mud filtrateand formation-water resistivity.

In the plains coals, the AIT analysis indicatedgreater invasion where the FMI tool displayedtensional fracturing (next page). The 1-ft [0.3-m]resolution measurement was able to resolve theeffects of invasion in the vicinity of a fault seen on

24 Oilfield Review

deg0 90

BeddingTrue Dip

0 120 240 360

XX92

XX93

Maximum horizontalstress direction Minimum horizontal

stress direction

Inducedfracture

Boreholebreakout

Dept

h, m


Orientation North

Boreholebreakout N45E

Drilling-inducedfractures S45E

> In-situ stress determination from borehole images. During drilling operations, stress release aroundthe borehole causes induced fractures and borehole breakout (left). These phenomena indicate thedirection of in-situ stresses. The orientations of these features, interpreted from FMI data (right), areused in hydraulic fracture treatment and deviated well designs.

API0 150

Gamma Ray

125 375mm

Caliper 2

125 375mm

Caliper 1

Bit Size

125 375mm

0 10


1 0m3/m3

Neutron Porosity

1000 2650kg/m3

Density

2e-05 cm 0.2

Fracture Aperture

0.2 2000ohm-m

0.2 2000ohm-m

120 240 3600

deg0 90

Resistive FractureTrue Dip

deg0 90

Drilling-Induced FractureTrue Dip

deg0 90

Conductive FractureTrue Dip

deg0 90

Bed BoundaryTrue Dip

XX20

XX22

XX24

XX18

XX16

Mea

sure

d de

pth,

m

Gamma Ray<45 API


Missing Core

Coal




Orientation North

> Montage of British Columbia foothills coal interval. The high degree of fracturing in the foothillscoals can make fullbore core recovery difficult. The interval shown was cored, but a short but crucialsection of core was lost from XX19 m to XX20 m. The FMI image, acquired across the interval, showedthat the missing core interval was heavily fractured. Gamma ray and caliper data are displayed inTrack 1 with borehole orientation. Track 2 contains porosity and lithology information. Fracture aper-tures calculated from FMI data are generally lower than in the plains coals and are displayed withresistivity data in Track 3. Track 4 contains the dynamic FMI image from which bedding and fractureplanes were picked. Track 5 shows the dip plots from the interpretation of Track 4.

13. Ali AHA, Brown T, Delgado R, Lee D, Plumb D, SmirnovN, Marsden R, Prado-Velarde E, Ramsey L, Spooner D,Stone T and Stouffer T: “Watching Rocks Change—Mechanical Earth Modeling,” Oilfield Review 15, no. 2(Summer 2003): 22–39.

Autumn 2003 25

API0 150

Gamma Ray

Bit Size

mm125 375

Caliper

mm125 375

900 -100

Density Correction

kg/m3

0 20


1 0m3/m3

Neutron Porosity

1000 2650kg/m3

Density

Poisson’s Ratio

ohm-m0.2 2000

Mea

sure

d de

pth,

m

VolumetricAnalysis

1 0vol/vol

Moved Water

Moved Hydrocarbon

Water

Gas

Quartz

Coal

Bound Water

Illite

Calcite

06000 6000






ohm-m0.2 2000

Flushed ZoneResistivity

ohm-m0.2 2000

ohm-m0.2 2000

ohm-m0.2 2000

ohm-m0.2 2000

0 0.5

0 100GPa

Static Young’sModulus

Outer InvasionDiameter

Inner InvasionDiameter

CoalGamma Ray

<45 API


mm

XX70

XX75

XX80

XX85

> Invasion analysis in the Alberta plains coals. Using a ramp-style invasion model and AIT ArrayInduction Imager Tool data, the plains coals show invasion up to 3.5 m [11.5 ft] in Track 4. Increasedinvasion is associated with intervals showing tensional fracturing on the FMI images. The 1-ft resolu-tion AIT measurement was able to resolve the effects of invasion near a fault seen on FMI images atXX79.5 m. Log analysts use this information to gauge the amount of invasion, which may be related toreservoir permeability. Track 1 displays gamma ray and caliper data. Track 2 contains porosity andlithology information, and Track 3 contains resistivity data. Track 4 shows the invasion calculation, andTrack 5 contains mechanical properties data, which show a higher Poisson’s ratio and lower Young’smodulus in the coals. Track 6 displays ELANPlus Elemental Log Analysis lithology results.

the FMI image at XX79.5 m. Further investigationis needed to establish correlations to producibil-ity. In contrast, the shear fractures observed onthe FMI images in the foothills coals were associ-ated with zones showing less invasion on the AITinvasion analysis (above). The log analystsbelieve this analysis provides a dependable way togauge the degree of invasion, which may correlateto reservoir-scale permeability.

Information from logs, core and outcrop canbe used in well construction. Proper cementingof Canadian CBM wells is a major challengebecause of the fractured state of coals.Frequently, primary cementing jobs fail to obtainor maintain cement returns to the surface,resulting in low cement tops and greater risk ofgas migration. Historically, operators have reliedon increasing the amount of excess cement

pumped to combat low cement tops, but a novelsolution known as CemNET advanced fibercement has yielded excellent results.

The CemNET slurry contains silica fibers thatbridge and plug lost-circulation areas, allowingthe slurry to return up the annulus. Operatorsare benefiting from this unique technology bypumping less cement, significantly reducingcement-disposal costs and potential damage tothe coals. The long-term benefit is bettercemented wells without remedial cementingcosts. In extremely problematic lost-circulationareas, CemNET fibers, coupled with theLiteCRETE slurry system, have proved successfulin CBM areas in Canada and in Wyoming, USA.14

The combination of these technologies inLiteCRETE CBM cement minimizes lost circula-tion problems, providing better cement coverage

that has helped reduce screenouts during fracture-stimulation treatments in some RockyMountain areas. Additionally, operators cancement a well with a single production-qualitycement back to surface, which no longer posesany constraints on the completion strategy.

Properly cemented wells prepare the way forsubsequent completion challenges. In CBM pro-duction areas worldwide, typically coal seamsfirst need to be dewatered to achieve maximumgas production. This is also true in Canada,although many dry coal seams have been found.When stimulating coals that have minimal waterin their cleat systems, or low-pressure coals, acompatible fracturing fluid system minimizesdamage to the permeability network. In Canada,fracturing fluid selections have included purenitrogen only, guar-based systems or the polymer-free ClearFRAC fracturing fluid.15 Thesefluids have been foamed using either nitrogen orcarbon dioxide. The move to polymer-free systems and foaming helps ensure improved fluid flow to the wellbore without damaging coal permeability.

Another common characteristic of CanadianCBM targets is that they consist of multiple thincoal seams; it is not unusual to have more than20 seams present. Schlumberger CoilFRAC stim-ulation through coiled tubing technology hasallowed operators to economically perforate andfracture all of these zones individually in a one-day operation.16 In some areas, Schlumberger isfracturing more than 30 zones per well and canstimulate two wells per day in certain circum-stances. Operators benefit from reduced setupcosts, reduced gas flaring and significantlyreduced time from completion to gas sales.CoilFRAC operations are suitable for environ-mentally sensitive areas since the equipment hasa smaller footprint than a service rig and most ofthe equipment travels to the lease only once.

Efforts to exploit Canada’s vast CBM resourceshave just begun. Armed with the historical knowl-edge of the coal mining industry, Canada’s CBMoperators continue to seek out optimal methodsfor drilling, evaluating, completing and producingcoalbed reservoirs.

Development in the Raton BasinThe Raton basin is located in the southern RockyMountains, along the boundary between NewMexico and Colorado, USA. It was formed duringthe late Cretaceous period and the early Tertiaryperiod. The Laramide uplift led to the erosion ofthe ancestral Rocky Mountains and the creationof an eastward prograding wedge of fluviodeltaicsedimentation, including the deposition ofnumerous coal beds. The basin contains two

26 Oilfield Review

API0 150

Gamma Ray

Bit Size

mm125 375

Caliper

mm125 375

900 -100

Density Correction

kg/m3

0 20


1 0m3/m3

Neutron Porosity

1000 2650kg/m3

Density

XX00

Poisson’s Ratio

ohm-m0.2 2000

Mea

sure

d de

pth,

m

VolumetricAnalysis

1 0vol/vol

Moved Water

Moved Hydrocarbon

Water

Gas

Quartz

Coal

Bound Water

Illite

Calcite

06000 6000






ohm-m0.2 2000


ohm-m0.2 2000

ohm-m0.2 2000

ohm-m0.2 2000

ohm-m0.2 2000

0 0.5

0 100GPa

Static Young’sModulus

Outer InvasionDiameter

Inner InvasionDiameter

CoalGamma Ray

<45 API


mm

XX05

> Invasion analysis in British Columbia foothills coals. The foothills coals show relatively low invasion,between 1 and 2 m [3 and 6 ft]. Shallow invasion profiles are observed in zones where the FMI imageshowed a high degree of shear fractures. Track 1 displays gamma ray and caliper data. Track 2 contains porosity and lithology information and Track 3 contains resistivity data. Track 4 shows theinvasion calculation, and Track 5 contains mechanical properties data, which show a higher Poisson’sratio and lower Young’s modulus in the coals. Track 6 displays ELANPlus lithology results.

Autumn 2003 27

coal reservoir systems: the primary productiontarget, the Vermejo formation coals, at an average depth of approximately 2000 ft [610 m],and the overlying Raton formation coals, the secondary coal target.

The Vermejo coals are moderately continuousbecause they were deposited in swamps and infloodplains within a fluvial-dominated deltaplain. Vermejo coals reach a combined thicknessof up to 40 ft [12 m] and average 20 ft [6 m] incombined thickness, with an average individualseam thickness of 2.6 ft [0.8 m], over a 275-ft [84-m] gross interval. By contrast, the Ratoncoals are thinner and less continuous becausetheir deposition was typically overbank into thebackswamp environments associated with mean-dering river systems. Raton coals can exceed75 ft [23 m] in gross thickness, but individualseams average 1.5 ft [0.5 m] in thickness.

During the Miocene epoch, an igneous complex called the Spanish Peaks intruded intothe basin.17 Igneous activity formed a complex network of dikes, sills and fractures that haveinfluenced the reservoir characteristics of bothcoals and sandstones (right).18 The mid-Tertiaryburial and the late-Tertiary uplift and erosion inthe southern part of the basin, coupled with thelate-Tertiary intrusions and associated heating,caused the overall fluid pressure in the basin todrop.19 This complicated geologic history has made the Raton basin difficult to understand and develop.

With CBM operations in several US basinsand over 1.9 Tcf [54.4 billion m3] in CBMreserves, El Paso Production Corporation hasstudied the Raton basin extensively since 1989.El Paso has drilled more than 350 wells and hasrecovered more than 42,000 ft [12,800 m] of full-bore core in the basin, making these coals some ofthe most studied CBM reservoirs in the industry.Vast amounts of lithologic, gas-content andisotherm data from cores taken across El Paso’sacreage have been examined and used to modelthe CBM reservoirs. These data have also beeninstrumental in the calibration of log interpreta-tion techniques, including ELANPlus Elemental

14. LiteCRETE cement is a unique system based on the prin-ciple of trimodal particle sizes. At low cement densities,it exhibits compressive strength similar to normal densitycements and maintains significantly lower permeabili-ties. For more on LiteCRETE slurry:Low N, Daccord G and Bedel J-P: “Designing FiberedCement Slurries for Lost Circulation Applications: CaseHistories,” paper SPE 84617, presented at the SPEAnnual Technical Conference and Exhibition, Denver,Colorado, USA, October 5–8, 2003.Junaidi E, Junaidi H, Abbas R and Malik BZ: “Fibers InCement Form Network to Cure Lost Circulation,” WorldOil (June 2003): 48–50.Walton D, Ward E, Frenzel T and Dearing H: “DrillingFluid and Cementing Improvements Reduced Per-FtDrilling Costs by 10%,” World Oil (April 2003): 39–47.

Ratonbasin

Walsenburg

Raton

ColoradoNew Mexico

Sier

ra G

rand

e A

rch

Basin axis

Alluvium, slopewash andlandslide materialBasalt flowsHuerfano formationMiddle Tertiary intrusivesCuchara formationPoison Canyon formationRaton formationVermejo formationTrinidad sandstone andPierre shale undividedPierre shale/Niobrara undividedPrecambrian rock undividedRaton basin boundary

Holocene andQuaternary

Tertiary

Cretaceous

15 miles0

0 20 km

ApishapaArch

Sang

rede

Cris

toM

ount

ains

Las

Anim

asAr

ch> Surface geology of the Raton basin. The 2200-square mile [5700-km2] basin contains two coal reser-voir systems: the primary production target, the Vermejo formation coals (pale yellow), at an averagedepth of 2000 ft [610 m], and the overlying Raton formation coals (light brown), which is a secondarycoal target. Tertiary igneous sills and dikes of the Spanish Peaks intrusion (magenta) have alteredcoals locally. (Adapted from Flores and Bader, reference 18.).

Al-Suwaidi A, Hun C, Bustillos J, Guillot D, Rondeau J,Vigneaux P, Helou H, Martínez Ramírez JA and ReséndizRobles JL: “Light as a Feather, Hard as a Rock,” OilfieldReview 13, no. 2 (Summer 2001): 2–15.

15. Chase B, Chmilowski W, Marcinew R, Mitchell C, Dang Y,Krauss K, Nelson E, Lantz T, Parham C and Plummer J:“Clear Fracturing Fluids for Increased Well Productivity,”Oilfield Review 9, no. 3 (Autumn 1997): 20–33.

16. Degenhardt KF, Stevenson J, Gale B, Gonzalez D, Hall S,Marsh J and Zemlak W: “Isolate and Stimulate IndividualPay Zones,” Oilfield Review 13, no. 3 (Autumn 2001):60–77.

17. Rose PR, Everett JR and Merin IS: “Potential Basin-Centered Gas Accumulation in Cretaceous TrinidadSandstone, Raton Basin, Colorado,” in Geology of TightGas Reservoirs, AAPG Special Publication. Tulsa,Oklahoma, USA: AAPG (1986): 111–128.

18. Flores RM and Bader LR: “A Summary of Tertiary CoalResources of the Raton Basin, Colorado and NewMexico,” in 1999 Resource Assessment of SelectedTertiary Coal Beds and Zones in the Northern RockyMountains and Great Plains Region, U.S. GeologicalSurvey Professional Paper 1625-A.

19. Stevens SH, Lombardi TE, Kelso BS and Coats JM: “AGeologic Assessment of Natural Gas from Coal Seams inthe Raton and Vermejo Formations, Raton Basin,” GasResearch Institute Topical Report, GRI 92/0345, ContractNo. 5091-214-2316, 1992.

20. Rautman CA, Cooper SP, Arnold BW, Basinski PM, Mroz TH and Lorenz JC: “Advantages and Limitations ofDifferent Methods for Assessing Natural Fractures in theRaton Basin of Colorado and New Mexico,” in AssessingNatural Fractures in the Raton Basin, June 2002.

Log Analysis computations. Since 2001, El Pasohas acquired Platform Express and ECS data on290 wells, and DSI and FMI data on strategicallylocated wells across the Vermejo Park Ranch.

Borehole images have been used along with out-crop and core data in a comprehensive effort tomodel the basin’s fracture systems.20

Even with this extensive database, the Ratonbasin remains a challenging area in which tooperate because of numerous complicating factors. First, gas-content values in the Vermejoand Raton formation coals vary across the basin,ranging from 50 to more than 400 scf/ton [1.56 to12.48 m3/tonne], on an in-situ basis. The deeperVermejo coals are typically gas-saturated andlend themselves to log-based interpretation tech-niques. However, selected shallower Vermejo andmany Raton formation coals are undersaturatedto varying degrees because they have beenaffected by the basin’s complex burial, thermal,pressure and hydrological history. As a result,variations in gas saturation relative to theisotherm complicate efforts to model the coals’productive potential and make log-based estimation of gas content and saturation profilesmore difficult.

Another complicating factor is that the hotintrusions locally altered the rank and the cleatand fracture permeability of the coals. The alter-ation of coal to a higher rank directly affects itsproductive potential. The intrusive bodies havechanged bituminous coal into higher rank coal,so the impact on gas content is inconsistent andnot yet predictable.

El Paso’s understanding of the reservoirs andthe basin as a whole has allowed the company toimprove its models and adopt strategies indrilling, completion, stimulation and productionthat maximize environmentally sound exploita-tion. For example, El Paso drills Raton basinCBM wells using air as the drilling fluid, therebyminimizing the damage to the coal’s cleat andnatural-fracturing systems. Wireline logging isaccomplished with air in the borehole by acquir-ing epithermal neutron data in combination withthe Platform Express tool.21

The Platform Express tool is designed to minimize the adverse borehole-rugosity effectson the density measurement commonly observedin coals and in air-filled boreholes. Detailedlithology of both the coals and the surroundinglow-permeability gas sandstone is computedusing the ECS tool, and SpectroLith andELANPlus processing. Log-based proximate analysis is also performed in the coals to deter-mine the percentages of volatile matter, fixedcarbon, moisture and ash, based on benchmark-ing to voluminous core data. From these percent-ages, coal rank and adsorbed gas volume can beestimated (above). In addition, the logs provide aqualitative estimate of the degree of cleating.

The DSI tool also provides El Paso with valu-able information on fractures and in-situ stressfields by measuring shear wave anisotropy.Anisotropy causes shear waves to split into twocomponents, one polarized along the direction of

28 Oilfield Review

XX00

XX50

Dept

h, ft

0 200API

Gamma Ray

Gamma Ray < 751 0

ECS Capture Hydrogen

Caliper

6 16in.

Caliper > Bit Size

Gas

ohm-m 20002





Flushed Zone Resistivity

Formation WaterResistivity


ohm-m 20002

ohm-m 20002

ohm-m 20002

ohm-m 20002

ohm-m 20002

1 1.7g/cm3

Density 1-in.

1 1.7g/cm3

Density 2-in.

0.3 0.05ft3/ft3

Epithermal NeutronPorosity

Effective Porosity

1 8


0.3 0.05ft3/ft3

Density Porosity

Crossover

Illite

Bound Water

Coal

Quartz

Pyrite

Calcite

Dolomite

Carbonate

Gas

Water

Irreducible Water

Moved Hydrocarbon

Moved Water

Irreducible Water

0.25 0

Irreducible WaterFlushed Zone

ft3/ft3

0.25 0ft3/ft3

Effective Porosity

0.25 0ft3/ft3

Volume Water

1 0ft3/ft3

Water Saturation

Water

Ash

Fixed Carbon

Volatiles

Moisture

Hydrocarbon

ft3/ft3

WaterSaturation

PartiallyCleated

PoorlyCleated

WellCleated

Integrated CoalFootage

Hydrocarbon

Water

0.01 10mD

Permeablility to Gas

0.01 10mD

Permeablility to Water

0.01 10mD

Intrinsic Permeability

Estimated GasMcf/day

Mcf/day 3000

Estimated Gas

0.14

0.06

0.130.08

0.07

0.06

0.100.110.11

0.12

0.35

0.35

0.350.35

0.350.35

0.350.35

0.35

0.35 0.440.48

0.57

0.65

0.640.41

1.001.00

0.44

0.94

33.50

27.50 143.49

161.07

> Characterizing coal and noncoal resources. With ECS Elemental Capture Spectroscopy and Platform Express data, anELANPlus analysis is computed. Lithology is presented in Track 4. Proximate (Track 5) and cleat analysis (Track 6) provideinformation on coal quality. Computed permeabilities are in Track 7 and estimated gas production is displayed in Track 8. ElPaso also uses the ELANPlus processing to calculate the reserves in the surrounding sandstones and siltstones.

Autumn 2003 29

maximum velocity, and the other along the direc-tion of minimum velocity. With two transmittersand two sets of receivers oriented perpendicularto one another, the DSI tool can measure boththe in-line waveforms from receivers oriented inthe same azimuth as the transmitter, andcrossline waveforms from receivers oriented 90°from the transmitter.22

During the DSI measurement, there is no wayto know how the signals are oriented withrespect to anisotropy. However, with both in-lineand crossline waveforms, it is possible to performa mathematical rotation to find the azimuth ofthe fast shear wave, and to determine the veloci-ties of both fast and slow shear waves. This rota-tion relies on the fact that the crosslinewaveforms should vanish when the measurementaxis is aligned with the anisotropy axis. The pro-cessing also computes the energy in the crosslinewaveforms as a percentage of the total waveformenergy. When the two axes are aligned, the resultis known as the minimum energy and is zero ifthe rotation model is correct. The maximumenergy is the energy at 90°. The differencebetween minimum and maximum is known asenergy anisotropy and is the principal measure ofanisotropy from DSI data.

The polytectonic history of the Raton basin hasintroduced other complications. For example,late-Tertiary changes in the regional stresses fromcompression to tension, thought to be caused byRio Grande rifting to the west, have major impli-cations for field development, especially in termsof well placement and stimulation practices. Priorto acquisition of key log data by El Paso, the Ratonbasin’s maximum principal stress direction wasbelieved to be east-west, consistent with a compressional basin model. FMI images and DSIanisotropy data have shown that the maximumprincipal stress direction is actually north-south(above right). This change has significant implica-tions for planning field development and wellstimulations (see “Refracturing Works,” page 38).Fracture stimulation will tend to propagate in thisnorth-south direction and, given an east-westLaramide-age open natural-fracture system, optimal drainage aspect ratios are anticipated. Asa result, where possible, development wells arenot positioned due north-south or east-west of oneanother; this maximizes ultimate drainage areasand gas recovery.

Currently, El Paso is assessing two differenthydraulic fracture stimulation treatments in theRaton basin. The first is a low-polymer boratefracturing fluid and higher concentrations ofproppant, delivered using coiled tubing and

straddle packers. This technique has been bene-ficial in wells where six to eight different coalbedlayers have been identified for stimulation. Thesepolymer-base fluids have been more successful inareas that initially produce large amounts ofwater, and where cleat- and fracture-system damage is of less concern. However, in areaswhere the coals initially produce low volumes ofwater, degrading the permeability to gas withinthe cleats and fractures is likely with polymer liquids. In these areas, El Paso is evaluating asecond technique of pumping foamed nitrogendown casing to hydraulically fracture the coalsand place smaller proppant concentrations.

The complexity and variability in the Ratonbasin make it extremely difficult to gauge thesuccess of fracture stimulation treatments inwell-performance terms. The search for the idealtreatment continues, but it is generally agreedthat more information is needed on hydraulicfracture propagation in and around coals.

Coalbed Completion StrategiesCoals often are adjacent to productive sands thathave dramatically different mechanical proper-ties. Coal has a higher Poisson’s ratio and a lowerYoung’s modulus than sand, so coals tend totransfer overburden stress laterally and maintainhigher fracture gradients. Cleating and naturalfracturing in coals create complex hydraulic fracturing scenarios that are extremely difficultto model.23

Gamma Ray<75 API

Hole Azimuth

Fast Shear Azimuth

Fast Shear Slowness

Slow Shear Slowness

Fast Shear Azimuth

Gamma Ray

Dept

h, ft

X050

X100

0 deg 360

0 deg 360

Tool Azimuth

Hole Diameter Quality

5 in. 20

0 API 150

-90 deg 90

NE 15

NE 4

NE 11

NE 13

NE 0

NE 18

NE 11

NE 23

NE 21

NE 20

NE 11

NE 32

NW 8

NW 2NW 5

NW 6

NW 7

NW 41NW 61

NW 17

Azimuth Uncertainty

950 50µs/ ft

950 50µs/ ft

0 100

MaximumCrossEnergy

EnergyDifference

MinimumCrossEnergy

0 100Anisotropy

>168-164-82-40-2

100 0

0 100

Anisotropy–Slowness

Anisotropy – Time

Slowness Time

< Understanding thestress fields. Anisotropydata from the DSI toolare used to computethe fast shear directionthat corresponds withthe maximum horizon-tal in-situ stress direc-tion. Here, the fastshear direction is ori-ented NNE to NNW(Track 2). The abruptshift in fast shearazimuth in the coal at adepth of X060 ft is notfully understood.

21. The epithermal neutron measurement is based on theslowing down of neutrons between a source and one ormore detectors that measure neutrons at the epithermallevel, where their energy is above that of the surround-ing matter. In air-filled boreholes, the lack of hydrogendramatically changes the thermal neutron populationnear the detectors, invalidating the response of a stan-dard thermal neutron log. The epithermal measurementis less affected by the borehole and by using an array ofback-shielded detectors, as in the APS AcceleratorPorosity Sonde device, can be calibrated to give poros-ity. Also, by measuring neutrons at the epithermal level,the effects of thermal neutron absorbers are avoided.

22. Armstrong P, Ireson D, Chmela B, Dodds K, Esmersoy C,Miller D, Hornby B, Sayers C, Schoenberg M, Leaney Sand Lynn H: “The Promise of Elastic Anisotropy,” OilfieldReview 6, no. 4 (October 1994): 36–47.

23. Olsen et al, reference 4.

Devices such as the DSI tool help determineaccurate in-situ stress magnitudes and directionsto improve hydraulic fracturing designs. In addi-tion, borehole images allow determination of thepreferred fracture plane, which reflects the current stress conditions at the wellbore. Thisinformation is used to devise perforation strategies that maximize the efficiency ofhydraulic fracturing operations by reducing near-wellbore tortuosity effects that lead to earlyscreenout.24 The relationship between coal cleatsand horizontal stresses is equally important andcan help explain CBM production variationsbetween wells and between production areas.

The effectiveness of hydraulically fracturingindividual coals has been debated because ofthese inherent complexities. Proppant volumesused in coal stimulations can be as high as12,000 lbm/ft [17,700 kg/m] of coal, but the effective hydraulic fracture half-lengths are dis-appointingly low—rarely documented over 200 ft[60 m]. Hydraulic fractures can grow out of zoneor develop into complex fracture networks withinthe coal, often damaging coal permeability whenpolymer-base fracturing fluids are used.25

Some experts believe that CBM reserveswould triple if fracturing coals were as effectiveas fracturing sandstones. Mechanical propertiesfrom DSI data show the stress contrast betweencoals and surrounding layers, enabling engi-neers to predict fracture height and improvestimulation treatments (left). In areas whereadjacent sandstones have productive potential,operators are reexamining their perforating andstimulation strategies in coals and sandstones. Atechnique called indirect vertical fracturing(IVF) initiates the fracture in the less-stressedsandstones above or below the coal to ensureadequate fracture propagation.26 In coals, thistechnique succeeds because the vertical permeability of coal is frequently greater than itshorizontal permeability, reducing the need for ahydraulic fracture to completely pass throughthe coal to effectively drain it. Another reason forthe success of this technique in coals is due tothe contrast in fracture gradient between thesurrounding clastic rocks and the coal. This difference helps ensure fracture connection withthe coal seam along the length of the hydraulicfracture. This technique was first demonstratedin the Fruitland coal and Picture Cliff sands of the San Juan basin in New Mexico and is currently being employed successfully in the central Rocky Mountains.

30 Oilfield Review

0 API 200

Gamma Ray

-80 mV 20

SpontaneousPotential

6 in. 16

Caliper

2 2000ohm-m

2 2000ohm-m

2 2000ohm-m

2 2000ohm-m

2 2000ohm-m

Closure Stress Averaged

Closure Stress

Fracture Pressure

Fracture Gradient

0 10

0 10

Young’s ModulusDynamic

0 10

980 2380psi

980 2380psi

psi

psi/ft

0 0.5

0 0.5

psi/ft

106 psi

106 psi

106 psi

X400

X350

Dept

h, ft

Stress in coalis higher than insurrounding layers

Shale

Bound Water

Coal

Sand

Hydrocarbon

Water





AIT Resistivity 10-in.Young’s Modulus

Static

Young’s ModulusDynamic

Pore PressureGradient

Poisson’s RatioStatic

Poisson’s RatioDynamic

Poisson’s RatioStatic

> Stress contrast. Coals are typically more stressed than surrounding rocks (blue arrows). This contrast inhibits fracture growth within the coals and promotes fracture growth in surrounding sandsand siltstones. Multiple fractures of limited length can also be created in the coals, causing damage tocoal permeability, slower dewatering and reduced gas production. Where adjacent sandstones haveproductive potential, a technique called indirect vertical fracturing (IVF) initiates the fracture in theless-stressed sands above or below the coal. This creates fractures of greater half-length, which contact and drain the coal more effectively. Gamma ray and caliper data are shown in Track 1 andresistivity data are displayed in Track 2. Lithology and volumetric information is shown in Track 3.Track 4 contains Young’s modulus and pore pressure gradient data and Track 5 displays closure stressand fracture pressure data zoned for input into hydraulic fracture design programs. Poisson’s ratiodata are presented in Track 6.

24. Almaguer J, Manrique J, Wickramasuriya S, Habbtar A,López-de-Cárdenas J, May D, McNally AC and Sulbarán A: “Orienting Perforations in the Right Direction,”Oilfield Review 14, no. 1 (Spring 2002): 16–31.Manrique JF, Poe BD Jr and England K: ProductionOptimization and Practical Reservoir Management ofCoal Bed Methane Reservoirs,” paper SPE 67315, pre-sented at the SPE Production Operations Symposium,Oklahoma City, Oklahoma, USA, March 26–29, 2001.

25. Palmer ID, Puri R and King GE: “Damage to CoalPermeability During Hydraulic Fracturing,” paper SPE21813, presented at the SPE Rocky Mountain RegionalMeeting and Low-Permeability Reservoirs Symposium,Denver, Colorado, USA, April 15–17, 1991.

26. Olsen et al, reference 4.27. Gas lock is a condition sometimes encountered in a

pumping well when dissolved gas, released from solutionduring the upstroke of the plunger, appears as free gasbetween the valves. On the downstroke, pressure inside

a barrel completely filled with gas may never reach thepressure needed to open the traveling valve. In theupstroke, the pressure inside the barrel never decreasesenough for the standing valve to open and allow liquid toenter the pump. Thus no fluid enters or leaves the pump,and the pump is locked. It does not cause equipmentfailure, but with a nonfunctional pump, the pumping system is useless. A decrease in pumping rate is accom-panied by an increase of bottomhole pressure (or fluidlevel in the annulus). In many cases of gas lock, thisincrease in bottomhole pressure can exceed the pressurein the barrel and liquid can enter through the standingvalve. After a few strokes, enough liquid enters the pumpto break the gas lock, and the pump functions normally.

28. Schwochow, reference 2.29. Albright J, Cassell B, Dangerfield J, Deflandre J-P,

Johnstad S and Withers R: “Seismic Surveillance forMonitoring Reservoir Changes,” Oilfield Review 6, no. 1(January 1994): 4–14.

Autumn 2003 31

Dewatering MethodsIn a majority of CBM wells, water production iscrucial to the gas-production process. Successfuldewatering requires uninterrupted pumpingoperations to decrease the bottomhole pressureso gas will desorb from the matrix and diffuseinto the cleat systems as quickly as possible.Pumping methods vary according to area liftrequirements and economics. Pumps must handle large volumes of water and be resistant tocoal fines, proppant damage and gas lock.27 Theserequirements have made progressing cavitypump deployment one of the more attractive liftmethods for CBM applications. The selection anddesign of an appropriate lift method often are notstraightforward and should focus on capacity,efficiency and dependability.

Schlumberger engineers and scientists at theAbingdon Technology Center and CambridgeResearch Center in England are developing soft-ware to aid in artificial-lift selection specific togas-well dewatering. The Gas Well DewateringSelection Tool (GDST) brings consistency to this critical selection process by utilizing theavailable well information to select the mostappropriate lift method. This software helpsSchlumberger field engineers, interacting withthe clients, use a selection process based onsound engineering practice. The tool provides acase-based reasoning engine and sensitivity analysis to obtain recommendations with defined confidence levels.

The economic drivers for CBM wells differfrom conventional gas wells in that most wellswill not require indefinite or increased dewater-ing through time. The GDST program enables the engineer to make several iterations to deter-mine the best lift method. The program does notprovide for comparative economics of lift meth-ods, although economic limitations of the pro-posed lift methods are considered in theselection process. The tool is designed to aid inthe selection of lift methods, including those thatmay not have been considered previously. (aboveright). An optimal dewatering strategy, coupledwith nondamaging cementing and stimulationtechniques, helps expedite water movement outof the coal’s fracture permeability network,thereby increasing well productivity.

Gas for the FutureThe exploitation of CBM resources is progressingsteadily. In the USA, natural gas prices havemade many areas—for example the Green Riverregion, Piceance basin, Arkoma basin andCherokee basin—more attractive for CBMdrilling, although some are not yet producing

significant volumes of natural gas. TremendousCBM reserves in the US Gulf Coast region haveyet to be tapped, but CBM activity has started inthe Cook Inlet, Alaska, USA.28 Worldwide, manycountries have just started investigating theirCBM resources. Local activity will grow out ofnecessity and out of the knowledge of how thesereservoirs behave.

Formation-evaluation methods, together withfullbore core data, are helping the industryunderstand coal reservoirs. Log processing tech-niques yield detailed lithology, and proximateand permeability data. Cleat and fracture systems are studied along with important local stress information through the use of bore-hole imaging techniques to more thoroughlyappreciate coal-seam permeability.

Coal-seam permeability, controlled by eventsthat occurred during deposition, maturation andtectonism, has surfaced as the most importantfactor in CBM production. Coal fracture systemsmust be connected successfully to the wellborethrough nondamaging stimulation methods.However, complex stress profiles and coal fracture systems make hydraulic fracture propa-gation in and around coals difficult to simulate.

New fracture-monitoring technology promisesreal-time images of hydraulic fracture creation.Early passive-seismic technologies performedprimitive hydraulic fracture monitoring, but

processing these data was tedious and time-consuming, and did not provide real-time information during fracturing operations.29 TheStimMAP hydraulic fracture stimulation diagnos-tics software allows real-time, onsite imaging ofhydraulic fracture seismic events, resulting inimproved job placement, enhanced well produc-tivity and a better understanding of fracturegeometry for future field-development decisions.

Although the industry’s knowledge of coal isvast and growing, modeling CBM reservoir behav-ior has been a challenging task. Schlumbergerhas improved coal reservoir modeling capabili-ties in ECLIPSE Office integrated simulationmanager and case builder software. This newsoftware incorporates isotherm data and handlesuncertainties, and will have the capability tomanage multiple gas types.

The nature of CBM development demandscareful economic consideration. Low-cost solu-tions can help, but technological advances indrilling, formation evaluation, completion, stimu-lation, production and reservoir modeling willhave a far greater impact. With immense world-wide reserves and a growing infrastructure toexploit them economically, coal ranks high onthe short list of unconventional fuels awaitingfuture development. —MG, JS

Reservoir Information

Production Information

Bottom Hole Flowing Pressure

Bottom Hole Static Pressure

Reservoir Temperature

Liquid Composition

Current Liquid Rate

Current Gas Rate (MMscfd)

Production Tubing Size (OD in inches)

SandProduction

Well Head Pressure

Sales Line Pressure

Well Depth

Casing Size (OD in inches)

Well Deviation

Requires Packer

Electricity Available

Injection/compressed Gas Available

Site Information

Comments

270 psi

psi1,000

F275

>30% Condensates

>600

Up to 2-7/8”

No

250 bbl/D

150

140

2,500

psi

psi

ft

>4-1/2”

High

Yes

No

No

Plunger Lift

Wellhead Compression

Velocity Strings

Siphon Strings

Foaming

Continuous Gas Lift

Intermittent GL Plunger

Intermittent Chamber Lift

Rod Pump

Hydraulic Jet Pump

ESP

PCP

Sufficient Information enteredConfidence

> The Gas Well Dewatering Selection Tool (GDST) software. The GDST helpsSchlumberger field engineers and clients select the most appropriate liftmethod, using a consistent selection process. Length of dark blue bars on theright indicates preferred dewatering methods.

producing natural gas from coal - schlumberger/media/files/resources/oilfield_review/ors03/... ·...

Documents