RESOURCE ARTICLE

Catching up with runaway hotplates$

In recent years, there have been numerous reports of “runaway hot plates”. This is to say, hot plates that heatuncontrollably regardless of the temperature setting or whether the controls are in the off position. Some ofthese events have resulted in injuries to laboratory personnel and damage to research facilities. Investigationsinto the cause of several of theseevents have determined that failure of a non-mechanical switch, a “triac”, in thehot plate can result in the circuit failing open, causing uncontrolled heating. The number of events in recentyears has led to greater awareness of the issue; however, in spite of this, devices utilizing this technologycontinue to be sold and used in research laboratories without additional controls to ensure their safety.

By Joseph M. Pickel,Mark Mathews,Kimberly Brown

INTRODUCTION

Hot plates are used for many applica-tions in research laboratories and canbe standalone devices (i.e., heatingonly) or can be a combination devicethat includes stirring functions. Theubiquitous nature of these devicescan lead to a casual approach withrespect to the hazards that they pres-ent. This complacency increases therisk of burn injuries and fires due tocontact of skin, combustible materials,or flammable chemicals with hot sur-faces. Many of these issues emerge dueto human error, poor work practices,and failure to consider these interac-tions when selecting a hot plate orsetting up an experiment. These issuesare common and are not discussed indetail in here. However, as a result ofevents that have damaged facilities ateach of the authors’ institutions, wehave become aware of a more insidiousissue associated with the design of hotplates—specifically that these platescan heat spontaneously and/or uncon-trollably under certain conditions. Inthis paper, we detail this problem andthe results of our investigations intothe extent of this issue and providerecommendations for the preventionof future occurrences.

On April 2nd, 2014 a BLEVE (Boil-ing Liquid Expanding Vapor Event)and fire destroyed a laboratory chemi-cal hood and caused significant dam-age to a laboratory at Oak RidgeNational Laboratory in Tennessee.Portions of the combination sash werejettisoned across the room, damaging a

flammable materials storage cabinet,and the heat of the fire burned a holeinto the epoxy surface of the hood. Thelab was unattended at the time of theevent, and the hood was empty exceptfor a combination hot plate and stirrer(aka stirring hotplate) and a closed250 mL flask containing hexane. Avacuum pump below the hood wasplugged in but turned off as was theheating controller on the hot plate (thestirring controller was in the “On”position). Although initial theoriesregarding the cause of the fire focusedon the vacuum pump, it was later con-firmed through forensic analysis thatthe hot plate had spontaneously anduncontrollably begun heating. Aninvestigation into the reason andmechanism for this phenomenon isnoted below.

HOT PLATE DESIGN AND OPERATION

In general, hot plates operate by apply-ing power to a resistive circuit attachedto a ceramic plate. The level of heatingis controlled by varying the power pro-vided to the resistor. Earlier designs forhot plates included bimetallic switchesthat controlled the duration of thepower delivery cycle or mechanicalswitches, which directly controlledthe amount of power provided to theheating element. More recent designshave incorporated a bidirectional orbilateral triode thyristor, more com-monly known as a TRIAC, which is asolid state switching device embeddedon a printed circuit board. This

Joseph M. Pickel is affiliated with OakRidge National Laboratory, P.O. Box2008,OakRidge,TN37831-6209,UnitedStates (e-mail: pickeljm@ornl.gov).

Mark Mathews is retired from Oak RidgeNational Laboratory, United States.

Kimberly Brown is affiliated with TheOffice of Environmental Health andRadiation Safety, University of Pennsyl-vania, 3160 Chestnut St., Suite 400, Phi-ladelphia, PA 19104-6287, United States(e-mail: kimibush@ehrs.upenn.edu).

$Notice: This manuscript has beenauthored by UT-Battelle, LLC, undercontract DE-AC05-00OR22725 withthe US Department of Energy(DOE). The US government retainsand the publisher, by accepting thearticle for publication, acknowledgesthat the US government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproducethe published form of this manuscript,or allow others to do so, for US gov-ernment purposes. DOE will providepublic access to these results of feder-ally sponsored research in accordancewith the DOE Public Access Plan(http://energy.gov/downloads/doe-public-access-plan).

4 ã 2018 Published by Elsevier Inc. on behalf of Division of Chemical Health and Safety of the American

Chemical Society.

1871-5532

https://doi.org/10.1016/j.jchas.2018.10.001

provides indirect control of the powerprovided to the heating element. Theuse of a TRIAC in the hot plate designprovides some improvement in safetyover the previous designs due toreduced spark risk and potential cor-rosion of the bimetallic thermostat.1

TRIACS also have the benefit of beingrelatively inexpensive to manufacturerin comparison to other technologies.Following the fire at its own organi-

zation, Oak Ridge National Labora-tory studied in detail the mechanismsby which hot plates incorporatingTRIAC switches operate and how theycould fail. In the paragraphs below andFigures 1–4, the electrical schematicsfor the normal operational and poten-tial failure modes for these devices areoutlined.

Mode 1 — Normal Condition (DeviceOff)

At the start of normal operations, thehot plate is plugged into a receptacle,but the heat control switch is turnedoff. The microprocessor does not turnon the solid state switching compo-nent, and, therefore, no powerreaches the hot plate heating element(Figure 3).

Mode 2 — Normal Condition (DeviceOn)

The heat control switch has beenturned on. It provides an input to themicroprocessor, which in turn sends acontrol signal to the switching compo-nent. The switching component con-trols the 120-V power to the heatingelement according to the micro-processor’s signal (Figure 4). The tem-perature feedback loop helps themicroprocessor keep the heating ele-ment at the desired temperature.

Mode 1 — Failure Condition (DeviceOn)

Unfortunately, a well-known failuremode of the switching component typ-ically used in hot plates is that it failsclosed (i.e., as a short circuit). Aswitching component in this conditionwill allow current to flow directly tothe heating element, no longerresponding to the microprocessor’scontrol signal. As a result, the hot platewill heat to its maximum temperature.The temperature feedback loop is of no

help in controlling the hot plate tem-perature. This hot plate has run away(Figure 5).

Mode 1 — Failure Condition (DeviceOff)

The heating control switch has beenturned off. But, the failed solid stateswitching component still passes 120-V power to the heating element, per-mitting the hot plate to stay hotbecause the microprocessor cannotcontrol the switching component(Figure 6).

EXTENT OF CONDITION

At the time of the Oak Ridge fire,researchers and others at the facilitywere not aware of the propensity ofcertain hot plates to heat spontaneouslyor uncontrollably, nor the mechanismfor this failure. Initial communicationswith other government laboratoriesindicated that the incident was isolated.However, after a second runaway hotplate was discovered (fortunately beforeany damage had occurred), efforts weremade to determine the extent of this

Figure 1. Infrared thermometer indicating elevated temperature on surface of hotplate while heating controls are in “OFF” position.

Figure 2. Damage to hood following BLEVE event.

Journal of Chemical Health & Safety, March/April 2019 5

condition in order to determine appro-priate corrective actions and to commu-nicate this risk to others.Information regarding the frequency

of events involving runaway hot plates

was obtained via internet searches ofpublicly available websites, discussionson the ACS Division of ChemicalHealth and Safety List-service, and asurvey conducted asking safety

professionals their experience withhis issue. Collectively, these sourcesrevealed that numerous research insti-tutions had experienced incidentsinvolving hot plates over a long period

Figure 3. Electrical schematic for hot plate plugged in during normal operations.

Figure 4. Electrical schematic for hot plate plugged in with power on during normal operations.

6 Journal of Chemical Health & Safety, March/April 2019

of time—some with significant damageas a result. A summary of hot plateevents that can be found on publiclyavailable websites is shown in Table 1.

Several of the event reports refer-enced in Table 1 provide detailedanalysis of the cause of defective hotplates consistent with that described

previously in this article, listings ofmodels susceptible to runaway events,and/or recommendations for preven-tion of recurrence. Some of these are

Figure 5. Electrical schematic for hot plate showing effect (uncontrolled heating) of failed solid state switching component.

Figure 6. Electrical schematic for hot plate showing effect (spontaneous uncontrolled heating) of failed solid state switchingcomponent when heating controls are in off position.

Journal of Chemical Health & Safety, March/April 2019 7

summarized later in this article. Infor-mation of this nature was alsoobtained from reports on other acces-sible websites15 but which did notidentify any specific events beyondthose listed in Table 1.In 2017, a survey was conducted by

members of the ACS Division ofChemical Health and Safety to solicitadditional information concerning hotplate failures and institutional failureprevention practices. This electronicsurvey was conducted using GoogleForms and distributed via email onthe ACS Division of Chemical Healthand Safety’s e-mail list (DCHAS-L16)as well as the Campus Safety Healthand Emergency ManagementAssociation’s community pages.17

The survey was conducted from April12 through 21, 2017 and 25 individualsresponded regarding their knowledgeof incidents involving hot plates.Twenty-one of the respondents to

this anonymous survey reported a totalof forty events involving hot plates,specifically citing: spontaneous heat-ing (13), runaway/uncontrolled heat-ing (19), or another unspecified mal-function (8). Respondents identifiedfive specific hot plate makes and mod-els involved in these events (see below)and identified equipment ages rangingfrom over 5 years old (4) to under 5years (6) with some hot plates ofunknown age (3).Like previous reports, the survey

identified several specific hot platemodels involved in the various eventsas shown in Table 2. Informationregarding damages incurred as a resultof these events was also obtained aspart of this survey, however they arenot repeated here as the range ofimpacts (no impact to significant dam-age) are similar those reported in Table1.In the time between the initiation of

this project and its publication, severaladditional events have been brought tothe attention of the authors via per-sonal communications or as obtainedfrom news reports. Given the pervasiveuse of hot plates in research laborato-ries, the number of events involvingfaulty hot plates far exceeds the limitednumber reported herein, and efforts tocommunicate this risk and preventrecurrence are warranted.T

able

1.Summary

ofEve

nts

InvolvingHotPlate

DescribedonPublicly

Available

Websites.

Institution(D

ate

ofRep

ort)

Typ

eofHotPlate

Involved

Cause

Iden

tified

Damage

Lawrence

Berkeley

NationalLaboratory

2

(June2005)

CorningPC

420

Hea

tingwhilein

offposition

Firein

fumehood

University

ofCalifornia,Santa

Cruz3

(April2007)

CorningPC-420

Hea

tingwhilein

offposition

Nospec

ifics-citesseveral“severalexplosionsor

fire

even

tsinvolvingdefec

tivehotplatesthat

resulted

ininjuries

tolaboratory

employe

esor

damage

tolaboratories”

Massach

usettsInstitute

ofTec

hnology

4

(July

2009,April2010,Feb

ruary

2011)

CorningPC-35

CorningPC-351

Notspec

ified

“Fires

occ

urred

”

University

ofCalifornia,Berkeley

Cruz5

(�June2011)

CorningPC-400D

Hea

tingwhilein

offposition

Disco

vered

priorto

adverse

impact

University

ofPen

nsylvania

6(M

arch

2011)

NotIndicated

Hea

terturned

oninstea

dofstirrer

Firein

fumehood

University

ofPen

nsylvania

7(M

ay2012)

NotIndicated

Hea

terlefton

Firein

laboratory

OakRidge

NationalLaboratory

8

(April2014)

Fisher

Isotemp

Hea

tingwhilein

offposition

BLEVE

damage

shoodandlaboratory

University

ofPen

nsylvania

9(July

2014)

CorningPC-320

Hea

tingwhilein

offposition

Firein

fumehood

OakRidge

NationalLaboratory

10

(January

2015)

Cim

erac

Hea

tingwhilein

offposition

Disco

vered

priorto

adverse

impact

University

ofPen

nsylvania

11(M

arch

2015)

CorningPC-420D

Hea

tingwhilein

offposition

University

ofPen

nsylvania

12(January

2016)

CorningPC-420D

Unco

ntrolled

(runawayhea

ting)

Smallfire

Emory

University

13(M

arch2016)

Fire—

Damage

tofumehood

NorthDakota

State

University

14

(Dec

ember

2017)

Notindicated

Hea

tingwhilein

offposition

Largefire

—sign

ifica

ntdamage

tolaboratory

andsm

okedamage

tofacility

8 Journal of Chemical Health & Safety, March/April 2019

MINIMIZING RISKS ASSOCIATEDWITH RUNAWAY HOT PLATES

Approacheshave beenproposedorpro-mulgated to prevent faulty hot plates. Inour research, the approaches identifiedfall into two categories—elimination ofthe risk by replacing the susceptible hotplate models with versions that providegreater safety controls, and administra-tively controlling the risk by institutingpolicies for the continued use of hotplatessusceptible to failure. Eliminationis the most effective way to protectworkers and facilities from this hazardand is recommended by the authors asthe preferred method for addressing thishazard.

ELIMINATE HEATINGFUNCTIONALITY

Many of the runaway/uncontrolledheating issues previously notedoccurred on dual functionality hotplates that provide both heating andstirring capabilities, some occurringwhen only stirring was required (i.e.,heat control in off position and stircontrol in the on position). It hasbeen observed that while stirringdevices without heating capabilitiesare often available, hot plates withstirring functionality are often uti-lized only for stirring out of conve-nience—to avoid requiring two sepa-rate devices for stirring and stirring/heating or to avoid having to switchout the devices between experiments.If only stirring is required, theauthors recommend that non-heatingplates be used.

ELIMINATION OF THE HAZARD VIAREPLACEMENT WITH INHERENTLYSAFER DESIGNS

Where possible, the most proactiveand conservative approach to is toreplace units having the failure-pronedesign with units that have inherentlysafer designs or alternative heatingmechanisms. Use of units with saferdesign eliminates the risk of runawayheating by inclusion of features such asautomatic over-temperature monitor-ing, on/off controls that are separatefrom the heating or stirring controls,and/or feedback loops. In addition,some designs reduce the issue ofhuman errors by having lights thatindicate when the surfaces are hot orincorporating hermetically sealedheating elements which eliminates anignition source if used near flammablevapors. Researchers interested inreplacing their existing hotplates withsafer models, should look for thesefeatures when considering replace-ment equipment.Some research institutions have

implemented voluntary or mandatoryhot plate replacement campaigns toremove error prone designs from use.While costly to implement, such pro-grams offer the research organizationsassurance that all hot plates in usemeet a minimum standard for safety.The implementation of such a cam-paign will necessarily vary with theorganization, however attributes ofsuccessful programs have includedproviding incentives to encouragereplacement, obtaining reduced pric-ing with vendors on preferred models,establishing a vendor or manufacturer

lead trade-in program, or a combina-tion of approaches.We note here that it is still possible to

obtain hot plates that have the error-prone TRIAC design (without addi-tional safety controls) so prospectivebuyers should confirm systems designsand safety features prior to making apurchase.

PREVENTING RUNAWAY HOTPLATES VIA ADMINISTRATIVEPRACTICES

In lieu of a program to eliminate failureprone hot plates, some institutions haveopted to implement programs and poli-cies to administratively control the haz-ard. While such administrative controlsare by no means as effective as eliminat-ing risk as engineering controls thateliminate the underlying problem, thesemay be selected as interim measuresuntil a means to fund and execute areplacement campaign is available.Although this can be accomplished ina variety of ways, common policiesinclude requiring attended operationsduring heating activities, and unplug-ging hot plates when not in use. It hasbeen noted that even when an institu-tion has made the effort to replace olderhot plates, administrative policies andoversight may still be required to someextent to ensure that runaway prone hotplates do not re-enter use (through per-sonnel moves or acquisition) and thatrequirements for proper use are dissem-inated to address issues not associatedwith equipment design.Instructions and markings on hot

plates listed to the UL 61010 series

Table 2. Hot Plate Models Involved in Events Identified by Survey Respondents.

Spontaneous Heating Events (Age of Plate in Years)a Runaway-Heating Events (Age of Plate in Years)a

� Corning PC 420 D (0.25–5)� Corning PC 320 (0.25–5)� Corning — model not specified (0.25-5)� Corning PC 35 (>5)� Corning PC 351 (>5)� Fisher brand — model not specified (>5)� VWR 7 � 7 with Aluminum Top (0.25–5)� Thermolyne SP46925 (>5)� Not Known/Identified (>5)

� Cimarec H-4954.xx (0.25–5)� Cimarec — two incidents, models not specified (0.25-5)� Corning PC 220 (>5)� Corning PC 420 (3–5)� Corning PC 420 D (<0.25)� Thermo Fisher — model not specified (unknown)� Commercial hot plate intended for home use (<0.25)

a Age of plate after purchase as reported by respondents.

Journal of Chemical Health & Safety, March/April 2019 9

of standards inform the user to unplugthe power cord when the hot plate isnot in use. Many organizations remindemployees to unplug hot plates whennot in use. In many cases, the plug isaccessible and can be removed fromthe supplying receptacle.Situations frequently arise in which

a hot plate’s plug is not readily acces-sible. A typical reason for the powerplug to be inaccessible is because thehot plate is inside a glove box, fumehood, or “hot cell” that has been sealedbecause the materials being studied orprocessed are particularly hazardousto persons, require a special atmo-sphere, or for similar reasons. Equip-ment (such as hot plates), glassware,chemicals, etc., are staged inside theglove box, fume hood or hot cell, andthen the box, hood or cell will be sealedby various methods for the duration ofthe experiment or process. Operationof the hot plate controls is accom-plished by utilizing gloves throughthe wall of a glove box, by remotemechanical manipulators, etc. In somecases, the receptacle supplying the hotplate is accessible after the equipmentand materials are loaded in the glovebox or fume hood and can be utilizedto disable the hot plate remotely. Butother times the glove box or fume hoodcontains materials in a configurationsuch that the receptacle cannot bereached and other (non-administra-tive) methods for controlling this haz-ard should be used.Installation of a switched receptacle

is a considered a good alternative tounplugging a hot plate, and this isencouraged when installing new gloveboxes and fume hoods. However, mod-ifying all of the existing glove boxes andfume hoods in use is expensive. In caseswhere older glove boxes or fume hoodsare used to routinely handle radioactiveor other hazardous materials, the cost ofretrofitting switched receptacles couldbe prohibitive.

FIXING THE PROBLEM

While unplugging a hot plate when notin use is effective, the control is faulty.This approach was also irrelevant insome of the cases noted in this paper

Hot plates, both combination andheat-only varieties, are an omnipresentpiece of equipment in most chemistrylaboratories. As an integral part ofmany research activities, they are usedwith and near a variety of potentiallyhazardous materials and environ-ments, and even when functioningnormally, can pose a hazard to theresearcher if not used properly. Mostresearchers are aware of these poten-tial hazards. However, we have foundthat many researchers are unawarethat hot plates with certain design fea-tures may heat unexpectedly oruncontrollably.

A multi-faceted approach is neededto remedy this issue and address poten-tial issues of hot plates already in use aswell as those still being manufacturedand sold.

- Research institutions and organiza-tions like the ACS Division of Chem-ical Health and Safety shouldimprove guidance and awarenessregarding the proper selection andsafe use of hot plates.

- Responsible manufacturers and dis-tributors should ensure that hotplates available on the market arenot susceptible to failure that couldcreate runaway or uncontrolledheating conditions.

- Trade and standards organizationsshould require that laboratory heat-ing devices have integrated safetyfeatures rather than relying onadministrative controls to preventequipment failures and theirconsequences.

Publicity surrounding events thathave caused injuries or significantdamage to research facilities hasencouraged many organizations totake action on the issue within theirown domains. Additional effort isneeded to increase awareness of thenumber of events that have occurredand the details surrounding the event.Actions of the American ChemicalSociety, the Pistoia Alliance, and Uni-versity of California Center for Labo-ratory Safety to develop lessonslearned databases to share this infor-mation can help with awareness.Researchers are encouraged to share

their experiences when adverse eventshappen to make these awareness pro-grams effective.While informed institutions and indi-

vidual researchers can take actions toeliminate or mitigate hazards associatedwith malfunctioning hot plates in theirown laboratories, substantive actionstaken by the manufacturers and distri-butors would have a significantly greaterimpact on educating the research com-munity and preventing the propagationof failure-prone devices in the market-place and the availability of devices withinherentlysaferdesigns.It isencouragingthat many vendors are actively identify-ing product design features that improvesafety, in addition to traditional focus onproduct specifications and performance.In addition, some vendors and distribu-tors have participated in buyback or hotplate replacement programs.18

The actions of responsible vendorsand distributors may be adverselyaffected if the initiatives are notadopted and sustained by the industryas a whole. The development or modi-fication of standards that require saferdesigns for hot plates could have asignificant impact in encouraging thisresult. To that end, in 2016 a requestwas made to the UL Standard Techni-cal Panel for Standard 61010 whichincludes laboratory heating devices(specifically in respect to UL61010-2-010: UL Standard for Safety ElectricalEquipment For Measurement, Control,and Laboratory Use) to consider add-ing a requirement to provide hot plateswith a mechanical switching device (aswitch, a contactor, etc.) that physi-cally disconnects ac power to the heat-ing element when the temperaturecontrol switch is turned to the “OFF”position. The issue was considered bythe responsible committee in 2017,however no action was taken by thecommittee as they believed the stan-dard adequately addressed the hazard.

CONCLUSION

In conducting research for this article,we have identified several reports ofrunaway hot plates, some with signifi-cant consequences. The failure of asolid-state “triac” component in some

10 Journal of Chemical Health & Safety, March/April 2019

hot plates has been identified as thecause of many of these events. Hotplates based on this design are still infrequent use in research laboratoriesand can still be purchased throughreputable vendors. Hot plates withinherently safer designs are availableon the market, and while some orga-nizations have made the effort to sys-tematically replace the failure-pronedesign with the safer technology, manyothers remain unaware of the potentialissue.Efforts are being made, through this

communication and through the workof organizations such as the ACS Divi-sion of Chemical Health and Safety, toincrease researchers’ awareness ofthis particular issue and to promulgaterecommendations for safe use of heat-ing devices. We note that manufac-turers, distributors and standardsorganizations have the opportunityto have a major positive impact onthe collective safety of their customersby eliminating the availability of hotplates susceptible to runaway heatingand by supporting product standardsthat require more reliable safetyfeatures.

REFERENCES1. http://www.labmanager.com/

lab-health-and-safety/2010/04/

10-tips-for-working-with-hot-plates,

accessed 5/3/2018.

2. http://www2.lbl.gov/ehs/Lessons/pdf/

FinalHotPlateLL.pdf, accessed 5/3/2018.

3. http://ehs.ucr.edu/laboratory/

hotplatesafetyadvisory20110715.pdf,

accessed 5/3/2018.

4. https://ehs.mit.edu/site/sites/default/files/Hot

%20Plates%20for%20Rep%20Mtgs%

20rev5%20Feb_2018.pdf, accessed 5/3/2018.

5. http://ehs.ucr.edu/laboratory/

hotplatesafetyadvisory20110715.pdf,

accessed 5/3/2018.

6. http://www.ehrs.upenn.edu/programs/

labsafety/alerts/, accessed 5/3/2018.

7. http://www.ehrs.upenn.edu/programs/

labsafety/alerts/, accessed 5/3/2018.

8. Event is the subject of this article.9. http://www.ehrs.upenn.edu/programs/

labsafety/alerts/, accessed 5/3/2018.

10. There are several reports about hotplate failures from different Depart-ment of Energy sites (including OakRidge National Laboratory) and uni-versities on DOE’s “OpExShare” site.To access information this information,please register on this page https://

opexshare.doe.gov/register.cfm, then use

the Search tool to search for “hot plate”.

11. http://www.ehrs.upenn.edu/programs/

labsafety/alerts/hot_plate_malfunction.

html, accessed 5/3/2018.

12. http://www.ehrs.upenn.edu/programs/

labsafety/alerts/hot_plate_malfunction.

html, accessed 5/3/2018.

13. https://scholarblogs.emory.edu/ranews/

2016/03/01/fire-safety-in-laboratories-

working-with-hot-plates/, accessed 5/3/

2018.

14. http://bismarcktribune.com/news/

state-and-regional/heating-device-may-

have-ignited-fire-at-ndsu-s-dunbar/article_

3bf7cf0f-b802-5553-92f3-41adfc7ee4fa.

html, accessed 5/3/2018.

15. https://www.drs.illinois.edu/News/

Warning-about-malfunctioning-hotplates,

accessed 5/3/2018.

16. Additional information regarding theDivision of Chemical Health andSafety List Service is available at thefollowing link: http://dchas.org/

the-dchas-l-list/.

17. Additional information regarding theCampus Safety, Health and Environ-mental Management Association com-munity pages is available at the follow-ing link: https://www.cshema.org/

communities.

18. https://ehs.mit.edu/site/laboratory-safety/

unsafe-hot-plates, accessed 10/15/2018.

Journal of Chemical Health & Safety, March/April 2019 11