HEATING TECHNIQUES- STATUS, IMPORTANT PARAMETERS- EXPERIENCES, OPTIMIZATION ANDDEVELOPMENT POSSIBILITIESGorm Heron, Vice President, ph.d.TerraTherm Inc.Incl. attachments:HEAT IT ALL THE WAY- MECHANISMS AND RESULTS ACHIEVED USING IN-SITU THERMALREMEDIATIONUSE OF THERMAL CONDUCTION HEATING FOR THE REMEDIATIONOF DNAPL IN FRACTURED BEDROCKGorm Heron, Vice President, ph.d.Ralph S. Baker, Chief Executive OfficerJohn M. Bierschenk, PresidentJohn C. LaChance, Senior Project ManagerTerraTherm Inc.ATV Jord og GrundvandAfværgeteknologier – State of the ArtSchæffergården, Gentofte22. oktober 2008
SUMMARYIn-situ thermal remediation has evolved significantly in the last 5 years, and it’s use is becomingmore and more wide-spread. To date, approximately 180 sites in North America and approximately20 sites in the rest of the world have been treated using one or more of the dominantthermal technologies:• Electrical Resistance Heating, also called Electro-Thermal Dynamic Stripping Process,• Steam Enhanced Extraction, also called Steam Injection, and• In-Situ Thermal Desorption, also called Thermal Conductive Heating.Each technology has it’s own niche and sweet-spot applications. The principle of the energydelivery is shown in Figure 1. The main mechanisms are thermal conduction for ISTD, resistive/ohmicheating for ERH, and steam injection and flow for SEE.All of the techniques are applicable for VOC contaminants located above the groundwatertable. Only ISTD is effective for SVOC contaminants that need treatment at temperaturesabove 120 o C. ERH and ISTD is challenged by high groundwater flow as cooling occurs, whereasSEE is applicable in such zones where the steam is readily injected /1/. Trends observedin the last decade include:• Thermal methods have been used to achieve better than 99% mass reduction, and mostoften is used only in source zones.• In the US, many thermal projects are guaranteed to meet the desired results. The confidencein the technologies is growing.• ERH is the most frequently used thermal technology. There are 3 vendors in the US;ISTD is represented by one, and SEE is offered by 2 vendors.• ISTD generally is used for sites with the most stringent goals (target soil concentrationsbelow 1 mg/kg).• SEE is used for deep and large sites without thick clay layers. The unit cost of SEE islower for such sites. There have been few such applications, but some large projects.• For VOC treatment of relatively large clays sites, both ERH and ISTD has been effectiveand reached unit treatment costs in the range of $100-200 per cubic meter. Smallsites have much higher unit costs.• Combinations of technologies have been used with great success only by a few vendors.When SEE is combined with either ERH or ISTD, treatment of both clay layersand high-flowing aquifers can be done simultaneously.• ISTD and SEE have or is being been used in Denmark at a total of approximately 7 sites.ERH has been used at a few sites in The Netherland and Belgium.In general, ISTD and ERH apply to the same VOC sites, except for those where only ISTDhas been effective:• Sites with very stringent remedial goals
• Sites with contaminants with high boiling points, requiring heating to above 120 o C• Fractured rock sites. An example ISTD application in saprolite and gneiss is providedin an attached paper /2/An accompanying paper presents the thermodynamic basis for thermal treatment, and someguidelines for implementation /1/. In the presentation, an overview of the heating methods,and a summary of results achieved, will be provided.Thermal Conduction Heating(TCH), or In-Situ ThermalDesorption* (ISTD)ComparisonElectrical Resistance Heating(ERH) – Joule HeatingSteam Enhanced Extraction(SEE) – Steam InjectionFigure 1. Illustration of the three major thermal technologies and how the energy is delivered. Brown layers representclays, the gray layer represents a permeable sand zone. Note that ERH for deep sites requires stacked electrodes,whereas ISTD uses a single heater pipe. Steam can flow readily in the sand, but not in the claysREFERENCES ATTACHED/1/ Heron, G., R.S. Baker, J.M. Bierschenk and J.C. LaChance. 2006. “Heat it All the Way - Mechanismsand Results Achieved using In-Situ Thermal Remediation.” Remediation of Chlorinated and RecalcitrantCompounds: Proceedings of the Fifth International Conference (May 22-25, 2006). Battelle, Columbus,OH./2/ Heron, G., R.S. Baker, J.M. Bierschenk and J.C. LaChance. 2008. “Use of Thermal Conduction Heatingfor the Remediation of DNAPL in Fractured Bedrock.” Remediation of Chlorinated and RecalcitrantCompounds: Proceedings of the Fifth International Conference (May 19-22, 2008). Battelle, Columbus,OH
HEAT IT ALL THE WAY - MECHANISMS AND RESULTS ACHIEVED USING IN-SITU THERMAL REMEDIATIONGorm Heron, TerraTherm, Inc., Keene, CA, USA)Ralph S. Baker, John M. Bierschenk, and John C. LaChance, TerraTherm, Inc., Fitchburg,MA, USA)ABSTRACTIn situ thermal remediation technologies have been proven to reach very low soil and groundwaterconcentrations by eliminating the dense non-aqueous phase liquid (DNAPL) source andreducing dissolved and adsorbed chlorinated volatile organic compound (CVOC) concentrationsto near non-detect levels. For chlorinated solvents, vaporization is the dominant mechanism,as vapor pressure and Henry’s law constants increase most markedly with temperature.For effective treatment, pneumatic and hydraulic control must be achieved during the heatingperiod, and a clear path for the generated vapors to an extraction system must be provided. Ifremedial goals are stringent, target temperatures shall be the in-situ boiling point of the soiland groundwater system, such that a phase change to the vapor state is forced by the heating.During operation, detailed temperature monitoring and process sampling is conducted andcompared to the performance calculated based on mass and energy balances. Interim and finalsediment sampling is used to verify remedial progress and performance prior to site demobilization.INTRODUCTIONRecently, results from sites that were heated and treated using in-situ thermal remediation(ISTR) have indicated impressive removal rates for DNAPL source zones in soil and groundwater.Published results from both a U.S. Department of Energy (DOE) site (Young-RaineySTAR Center; Heron et al. 2005), and an industrial facility in the Midwest (LaChance et al.2004) have documented mass removal efficiencies in the 99.9% range. Soil concentrations ofContaminants of Concern (COCs) below or near non-detect are reported, and groundwaterconcentrations near or below Maximum Concentration Limits (MCLs) have been observedinside the original source zones. These results appear almost unrealistic, considering the recalcitrantnature of DNAPLs in the subsurface, long-term diffusion processes, heterogeneityof most source zones, and frequently raised questions about DNAPL capture at thermal sites.Other site reports, particularly from sites where Electrical Resistance Heating (ERH) wasused, have reported less impressive results, sometimes less than 90% mass removal (LowryLandfill, CO: Plaehn et al. 2004; ICN Pharmaceuticals, Portland, OR: USEPA, 2004; NavyBedford, MA: Francis and Wolf, 2004). As more data are emerging, it is becoming evidentthat thermal remediation spans a wide range of heating methods, and that applications varyfrom very robust, effective systems to poorly designed and ineffective systems.This paper reviews the mechanisms behind ISTR critically, focusing on what happens at thepore and micro-scale, as well as larger scale during heating. It will review ways for the contaminantsto be mobilized and extracted, and show the most proper design for vapor recoveryand capture systems. Several key design elements will be presented, in an attempt to
explain why not all thermal projects have achieved the success made possible by the theory,and how achievement of desired results can be much more widely attained.THERMAL REMEDIATION METHODSISTR is gaining acceptance for restoration of NAPL source zones (Davis, 1997). The followingISTR methods are discussed below:• Steam Enhanced Extraction (SEE).• Electrical Resistance Heating (ERH).• Dynamic Underground Stripping (DUS).• In-Situ Thermal Desorption (ISTD).Steam has been used to heat the more permeable zones, which are typically sandy layers withrelatively low clay and mineral contents. The in-situ process using steam injection and aggressivefluids extraction was named Steam Enhanced Extraction (Udell et al. 1991), and severalfield demonstrations and full-scale cleanups have been conducted (Udell et al. 1999;Eaker 2003; EarthTech and SteamTech 2003). Mechanisms used in SEE were reviewed critically(Udell 1996).Both three-phase and six-phase ERH were developed as robust techniques in the 1990’s anddemonstrated in the field. ERH involves passing electricity through the soil between electrodes,and heating the soil by Joule heating. Laboratory studies demonstrated that thermodynamicchanges induced by ERH can lead to very effective removal of chlorinated solventsfrom silts and clays (Heron et al. 1998). Since the late 90’s, several commercial full-scale implementationsof both three- and six-phase ERH were completed, some by the trade nameElectro-Thermal Dynamic Stripping Process (ET-DSP) (McGee 2003).The combination of steam and ERH is named Dynamic Underground Stripping, and wasdemonstrated at a gasoline spill that had resulted in LNAPL contamination above and below arising water table at the Livermore Gas Pad (Newmark 1994; Daily et al. 1995). This methodwas used recently to remediate a DNAPL source area at the Young-Rainey STAR Center (Heronet al. 2005).A robust method for heating soils and groundwater is thermal conduction heating, also namedIn-Situ Thermal Desorption (ISTD; Stegemeier and Vinegar 2001). ISTD is a soil remediationtechnology in which heat and vacuum are applied simultaneously. Heat flows into the soilprimarily by conduction from heaters typically operated between 500 and 800°C, with the targetsoil volume being heated to 100°C for VOC removal. As the soil is heated, water is boiledand DNAPL constituents in the soil are vaporized. The resulting steam and vapors are drawntoward extraction wells for in-situ and aboveground treatment. Compared to fluid injectionprocesses, the conductive heating process is very uniform in its vertical and horizontal sweep.
Other thermal methods such as hot water flooding, hot air sparging, and radio-frequency heatingwere not consider during this review. This review focuses on volatile compounds, andtherefore on the methods for which heating to the boiling point of water is sufficient. Thehigher temperature version of TerraTherm’s ISTD technology, used for treatment of semivolatileorganic compounds (SVOCs) such as polychlorinated biphenyls (PCB) and coal tar,will not be discussed.REMEDIATION MECHANISMS FOR CVOC CONTAMINANTSFor thermal treatment of VOC DNAPL, the dominant mechanism is vaporization, as illustratedin Figure 1, showing how boiling leads to steam formation and gas flow rich in contaminantvapors out of the pore matrix. Note the continuous gas phase in the right image wherepore fluids are boiling and creating steam, which sweeps out to recovery wells. Boiling occursat DNAPL-water interfaces and throughout.1 mmFigure 1. Conceptual illustration of the difference between ambient temperature (left) and boilingtemperature conditions (right) at the pore scaleFigure 2 summarizes the physical property changes occurring during heating for water, trichloroethene(TCE), and tetrachloroethene (PCE). While DNAPL density, viscosity, surfacetension, and solubility varies slightly, vapor pressure and Henry’s law constants increase dramaticallywith temperature.
Density (g/mL)22.214.171.124.21.00.80.60.40.20.00 20 40 60 80 100o CViscosity (mPa s)2.01.51.00.50.00 20 40 60 80 100o CVapor pressure (atm)126.96.36.199.00.80.60.40.20.00 20 40 60 80 100o CSurface tension (mN/m)80706050403020100Interfacial tension PCE-waterSolubility (mg/L)240020001600120080040000 20 40 60 80 1000 20 40 60 80 100o Co CWaterTCEPCEHenry's law constant ( - )765432100 20 40 60 80 100o CFigure 2. Properties of water, PCE and TCE as a function of temperature.Other mechanisms include enhanced dissolution, hydrolysis, and aqueous phase oxidation.However, vaporization is dominant for most chlorinated solvents.HOW CLEAN CAN IT GET?Table 1 shows the results achieved using DUS combined with detailed flexible monitoring,sampling, energy balance calculations, and careful pneumatic and hydraulic control. Removalefficacies of 99.9% or better were achieved.Table 1. Treatment efficiency based on mass estimates from soil sampling before and after DUS treatment at theYoung-Rainey STAR Center, Largo, FL (Heron et al. 2005).Numberof samplesTCE(µg/kg)cis-1,2-Dichloroethene(µg/kg)MethyleneChloride(µg/kg)Toluene(µg/kg)Before, maximum231250,000 68,000 650,000 72,000Before, average2,753 1,239 3,444 825After, maximum 110 120 8.2 42080After, average3.3 4.4 0.8 21Removal efficiency (%) 99.99 99.85 99.99 99.89
Table 2 shows similar contaminant mass reduction achieved using ISTD to heat a tight saturatedclay to 100 ° C and vaporizing 40% of the groundwater.Table 2. Comparison of COC concentrations before and after ISTD treatment at the Terminal 1 site in Richmond,CA (Geomatrix and TerraTherm 2005).PCE(µg/kg)TCE(µg/kg)Numberof samplescis-1,2-Dichloroethene(µg/kg)Vinyl chloride(µg/kg)Before, maximum64510,000 6,500 57,000 6,500Before, average34,222 1,055 6,650 932After, maximum 44 < rep.limit 1,500 2444After, average12 < rep.limit 65 4.7Removal efficiency (%) 99.96 99.63 99.03 99.49These results illustrate the dramatic reduction in source zone CVOC concentrations when thetechnology is used to reach boiling point temperatures and operated with a sufficient energyinput to vaporize more than 30% of the soil moisture.CRITICAL DESIGN AND IMPLEMENTATION ISSUESSelect the Most Appropriate Heating TechnologyOften site owners or consultants fail to consider all the options for thermal treatment and relyon one concept and cost when making a decision. It is important to realize, for instance, thatERH and ISTD often apply to the same sites, and that sometimes one method is more appropriate.Key issues for the selection includes treatment size (ERH often is cheapest for smallsites with modest remedial goals), treatment depth (ISTD is simpler and more cost-effectivefor deep sites), treatment goals (ISTD is often used for sites with very stringent cleanupgoals), and site permeability (SEE may be more applicable to deep permeable formations).Therefore, clients and consultants should not restrict themselves to working with a singlethermal vendor, but ought to request evaluations from more than one for each site.Establish Hydraulic and Pneumatic ControlBefore heating, inward flow of fluids (vapor and water) must be ensured. During SEE andDUS, inward flow of groundwater and vapor must also be ensured. Detailed calculation ofnecessary rates must be performed (steam can rapidly displace large quantities of fluids). DuringERH and ISTD, vapor capture must be ensured by a robust vacuum extraction system thatallows a flow pathway for the steam and CVOC vapors to extraction points. One cannot relyon steam bubbles migrating upward to the vadose zone, or on clay drying to create permeability– the generated vapor must flow readily to extraction wells screened in the right zones.During ERH and ISTD, hydraulic control must be maintained, either by boiling sufficientamounts of groundwater to create capture, pumping, or through use of a hydraulic barrier.Heat to Target Temperature High Enough to Accomplish the Remedial ObjectivesFor most CVOCs, except those which degrade readily by hydrolysis such as Methylene Chlo-
ide (MeCl 2 ), 1,2-Dichloroethane (1,2-DCA), and 1,1,1-Trichloroethane (1,1,1-TCA), the targettreatment temperature should be the boiling point of the pore water or groundwater. Heatingto 100 o C or slightly lower where vacuum is applied ensures that all DNAPL is vaporizedand removed, and that steam stripping will reduce dissolved and adsorbed concentrations(Udell, 1996). One mistake is to rely on the psychrometric effect (air mixing with steam andlowering the boiling point), since typical heterogeneity makes it impossible to distinguishwhether a zone heated to 85 o C is being cleaned due to air flow or whether it is stagnant andby-passed by the heating (Heron et al. 2005). Ultimately this reduces the certainty of reachingcleanup goals in a timely manner.Below the water table the target temperature increases with depth due to the increased pressure.At a depth of 10 m below the water table, the groundwater boils at approximately 120 o C.Target temperatures must be the boiling temperature to ensure steam stripping.During operation, subsurface temperature monitoring is essential (e.g. Heron et al. 2005). Forheterogeneous sites, thermocouples should be placed no more than 1.5 m apart vertically, anda network of monitoring wells should cover the target area and the zones around it.Mass and Energy Balance Calculations and Data ManagementA key to a good thermal project design and execution is careful energy management andmonitoring of the progress in different ways – an example of the use of a detailed energy balancefor finding problems is presented in Figure 3. At this site, the energy balance indicatedthat a cool zone was present (average site temperature was 80-90 o C, but the thermocouplesshowed 100-110 o C), and subsequent drilling revealed a recalcitrant zone, which was then targetedfor more intense treatment. The results were encouraging, showing the importance ofcomplete heating, and the value of careful monitoring and engineering checks.Use Pressure Cycling during Steam Enhanced ExtractionDuring SEE and DUS, steam breaks through to dedicated extraction wells (Udell et al. 1991).The principle of pressure cycling is illustrated in Figure 4. The steam stripping is enhancedduring the depressurization step, where the pressure in the steam zone is reduced. This leavesthe pore fluids in the steam zone and the surrounding condensation zones at slightly superheatedconditions since the equilibrium temperature is lower than the actual temperature (seethe inserted steam pressure curves at the top of the figure). The zones respond by releasingenergy to get to a lower equilibrium temperature – this happens by boiling pore fluids. Thegenerated steam and BTEX-vapors migrate in the steam zone towards recovery wells. Thiswas documented to lead to large increases in vapor-phase recovery during full-scale remediation(Heron et al 2005). Failure to conduct pressure cycling will prolong the heating periodand reduce treatment efficiency.
Cumulative energy (million BTU)12,00010,0008,0006,0004,0002,000Steam inETDSPTotal injectedTotal energy extractedNet energy addition150125100755025Calculated average temperature (oC)0010/1 10/11 10/21 10/31 11/10 11/20 11/30 12/10 12/20 12/30 1/9 1/19 1/29 2/8 2/18 2/28Figure 3. Energy balance for the steam and ERH project at Young-Rainey STAR Center Area A. The bluecurve is calculated average treatment zone temperature (Heron et al. 2005).Figure 4. Illustration of pressure cycling during steam injection and extraction. The left figure illustratesthe pressurization and heating stage, the right the situation during de-pressurization.Use Process Stream Sampling to Document Mass Removal Rates over Time
A flexible approach to sampling and data collection can not only enhance the operation of thethermal system, it can shorten the time to reach remedial goals and lower overall costs. Extractionwells and manifolds are hot and under vacuum – no standard EPA methods exist forcollection and sample analysis. However, screening-level sampling can reveal areas whereproblematic compounds persist and reveal area where treatment is complete (diminishing returns).Examples are published by US DOE (2003).Use Interim Soil Sampling to Document Remedial ProgressInterim drilling during thermal remediation can be performed safely when not drilling intosignificant steam zones (Gaberell et al. 2002). The data can be used to document remedialprogress and final performance before operation is completed. It is useful for gathering informationin zones where contaminant extraction continues despite a sense that heating is complete,as shown by Heron et al. (2005).Soil and sediment sampling is particularly useful since potential rebound after cooling wouldbe the result if partitioning of contaminants from the sediments to groundwater. Since thesampling represents the potential source for rebound, it is preferred over waiting for coolingand then performing groundwater sampling.CONCLUSIONSThis review revealed how effective thermal remediation can be for CVOC sites if designedand implemented well. It identified these critical design elements:• CVOC sites with stringent treatment goals must be heated to the boiling point of water,and a significant amount of energy spent on boiling groundwater.• Pneumatic and hydraulic control must be maintained, and a path for the generatedsteam to extraction wells must be included.• Pressure cycling is important on steam projects – it shortens remediation time and increaseseffectiveness.• Mass and energy balance calculations are useful for overall process verification, troubleshooting,and assurance that site heterogeneities are addressed.• Interim process and soil sampling are important for tracking remedial performance andverifying treatment prior to cessation of operations.Remedial standards of below 1 mg/kg have been consistently met by properly engineeredISTR systems.REFERENCESDaily, W.D., A.L. Ramirez, R.L. Newmark, K.S. Udell, H.M. Buettner, and R.D. Aines. 1995. Dynamic UndergroundStripping: Steam and electric heating for in situ decontamination of soils and groundwater. US Patent #5,449,251.Earth Tech and SteamTech. 2003. Environmental Restoration Program, Site 61 treatability study report, Steaminjection, Northwest Main Base, Operable Unit 8, Edwards Air Force Base, California, September.
Eaker, Craig 2003. Southern California Edison Company, Visalia Pole Yard, Visalia, California, Draft paper inpreparation for submission.Davis, E.L. 1997. How heat can accelerate in-situ soil and aquifer remediation: important chemical propertiesand guidance on choosing the appropriate technique. US EPA Issue paper EPA/540/S-97/502.Francis, J. and J. Wolf. 2004. “In Situ Remediation of Chlorinated VOCs and BTEX Using Electrical ResistanceHeating.” Paper 2B-19, in: A.R. Gavaskar and A.S.C. Chen (Eds.), Remediation of Chlorinated and RecalcitrantCompounds —2004. Battelle Press, Columbus, OH.Gaberell, M., A. Gavaskar, E. Drescher, J. Sminchak, L. Cumming, W-S. Yoon, and S. De Silva. 2002. SoilCore Characterization Strategy at DNAPL Sites Subjected to Strong Thermal or Chemical Remediation. Paper1E-07, in: A.R. Gavaskar and A.S.C. Chen (Eds.), Remediation of Chlorinated and Recalcitrant Compounds.Geomatrix and TerraTherm, 2005. Final report for ISTD treatment at Terminal 1, City of Richmond. GeomatrixConsultants, Oakland, CA.Heron, G., M. van Zutphen, M.; T.H. Christensen, and C.G. Enfield. 1998. Soil heating for enhanced remediationof chlorinated solvents: A laboratory study on resistive heating and vapor extraction in a silty, lowpermeablesoil contaminated with trichloroethylene. Environmental Science and Technology, 32 (10), 1474-1481.Heron, G. S. Carroll, and S.G.D. Nielsen. 2005. Full-Scale Removal of DNAPL Constituents using Steam EnhancedExtraction and Electrical Resistance Heating. Ground Water Monitoring and Remediation, 25 (4), Winter2005, pp. 92-107.LaChance, J.C., R.S. Baker, J.P. Galligan, and J.M. Bierschenk. 2004. Application of Thermal Conductive Heating/In-SituThermal Desorption (ISTD) to the Remediation of Chlorinated Volatile Organic Compounds in Saturatedand Unsaturated Settings. Proceedings of Battelle’s Conference on Remediation of Chlorinated and RecalcitrantCompounds, Monterey, CA, May 24.McGee, B.C.W. 2003. Electro-Thermal Dynamic Stripping Process for in situ remediation under an occupiedapartment building. Remediation, Summer: 67-79.Newmark, R.L. (ed.) 1994. Demonstration of Dynamic Underground Stripping at the LLNL Gasoline Spill Site.Final Report UCRL-ID-116964, Vol. 1-4. Lawrence Livermore National Laboratory, Livermore, California.Plaehn, D.R., T. Powell, G. Beyke, S. Richtel, D. Bollmann and J. Herzog. 2004. “Remediation of a Waste PitUsing Electrical Resistance Heating.” Paper 2B-13, in: A.R. Gavaskar and A.S.C. Chen (Eds.), Remediation ofChlorinated and Recalcitrant Compounds —2004. Battelle Press, Columbus, OH.Stegemeier, G.L., and Vinegar, H.J. 2001. Thermal Conduction Heating for In-Situ Thermal Desorption of Soils.Ch. 4.6, pp. 1-37. In Chang H. Oh (ed.), Hazardous and Radioactive Waste Treatment Technol. Handbook, CRCPress, Boca Raton, FL.Udell, K.S., N. Sitar, J.R. Hunt, and L.D. Stewart. 1991. Process for In Situ Decontamination of Subsurface Soiland Groundwater. US Patent # 5,018,576.Udell, K.S. 1996. Heat and mass transfer in clean-up of underground toxic wastes. In Annual Reviews of HeatTransfer, Vol. 7, Chang-Lin Tien, Ed.; Begell House, Inc.: New York, Wallingford, UK, pp. 333-405.Udell et al. 1999. Alameda Point Site 5 Steam Enhanced Extraction Demonstration. Draft Final Report submittedto US Navy. Berkeley, CA.US DOE (2003). Pinellas Environmental Restoration Project. Northeast Site Area A NAPL Remediation FinalReport. Young - Rainey STAR Center. U.S. Department of Energy, Grand Junction Office, Grand Junction,Colorado. September.U.S. EPA. 2004. “Cost and Performance Report, Electric Resistive Heating at the ICN Pharmaceutical Site, Portland,OR. February 2004.” In: In-Situ Thermal Treatment of Chlorinated Solvents: Fundamentals and Field Applications.Office of Solid Waste and Emergency Response, Office of Superfund Remediation and TechnologyInnovation. EPA 542-R-04-010. March.
USE OF THERMAL CONDUCTION HEATING FOR THE REMEDIATION OFDNAPL IN FRACTURED BEDROCKGorm Heron, TerraTherm, Inc. Keene, CA, USARalph S. Baker, John M. Bierschenk, and John C. LaChance, TerraTherm, Inc. Fitchburg,MA, USAABSTRACTThis paper presents the first full-scale remediation at a fractured rock site using Thermal ConductionHeating (TCH), also known as In-Situ Thermal Desorption (ISTD). A 90-ft deep TCEsource area was treated thermally, including thick zones of saprolite and gneiss bedrock. Thethermal treatment used 24 heater borings/wells, and operated for 148 days, after which an averagetemperature of approximately 100 o C was achieved. The ISTD remediation work washighly successful at reducing soil, rock and groundwater concentrations at this confidentialfacility. Post remediation soil sampling indicated that the 95% UCL of the mean concentrationof TCE in soil within the treated area was 17 μg/kg. This was significantly lower thanthe remedial goal of 60 μg/kg. In addition, groundwater concentrations within the treatmentzone were reduced by between 74.5% and 99.7%. The total mass of VOCs removed from thesubsurface during the ISTD remediation was approximately 12,000 lbs, almost all of whichwas TCE.INTRODUCTIONPrior to this project, there did not exist an effective technology for the remediation of DNAPLin fractured bedrock systems. This is because DNAPL in fractured bedrock presents severalsignificant challenges; including: 1) defining the area to be treated; 2) potential impacts ofmatrix diffusion within and downgradient of the source zone; 3) discrete nature of fracturepathways and presence of dead-ends; and 4) accessing DNAPL within the fractures and thecontaminant mass in the matrix. One technology however, that may be able to overcomemany of these limitations is Thermal Conduction Heating also known as In-Situ Thermal Desorption.ISTD is the simultaneous application of heat, by TCH, and vacuum to the subsurfaceto remove organic chemicals. Heat is applied by installing electrically powered heaters atregular intervals throughout the zone to be treated. The heat moves out into the inter-well regionsprimarily by thermal conduction. Thermal conduction heating of fractured bedrocksites is capable of: 1) achieving thorough heating of the bedrock (matrix and fractures), 2)preventing unwanted condensation of steam and CVOC vapors, and 3) capture and removal ofthe CVOC mass liberated from the bedrock and unconsolidated deposits.In fractured rock settings, a substantial fraction of the contaminant mass may be located in therock matrix, dissolved in matrix porewater, adsorbed to mineral surfaces and organic matter,or even as tiny droplets or ganglia if the DNAPL has entered the matrix. As a result, it is notsufficient to only apply a remedy to the fracture systems, since back-diffusion and transport ofcontaminants back out of the matrix can make it impossible to achieve satisfactory plumeconcentration reductions. Therefore, an effective fractured rock remedy must involve treatmentof the contaminants in the matrix.
For thermal treatment of VOC DNAPL and dissolved and adsorbed phases, the dominant removalmechanism is vaporization, as illustrated in Figure 1 for an equivalent porous mediumshowing how boiling leads to steam formation and gas flow rich in contaminant vapors out ofthe pore matrix. Note the continuous gas phase in the right image where pore fluids are boilingand creating steam, which sweeps out to recovery wells. Boiling occurs at DNAPL-waterinterfaces and throughout. This mechanism has led to very effective thermal treatment even ofthick saturated clay layers (Geomatrix and TerraTherm 2005, LaChance et al. 2004).1 mmFigure 1. Conceptual illustration of the difference between ambient temperature (left) and boiling temperatureconditions (right) at the pore scale for a porous medium.Figure 2 summarizes the physical property changes occurring during heating for water, trichloroethene(TCE), and tetrachloroethene (PCE). While DNAPL density, viscosity, surfacetension, and solubility varies slightly, vapor pressure and Henry’s law constants increase dramaticallywith temperature (Heron et al 2006).
Density (g/mL)188.8.131.52.21.00.80.60.40.20.00 20 40 60 80 100o CViscosity (mPa s)2.01.51.00.50.00 20 40 60 80 100o CVapor pressure (atm)184.108.40.206.00.80.60.40.20.00 20 40 60 80 100o CSurface tension (mN/m)80706050403020100Interfacial tension PCE-waterSolubility (mg/L)240020001600120080040000 20 40 60 80 1000 20 40 60 80 100o Co CWaterTCEPCEHenry's law constant ( - )765432100 20 40 60 80 100o CFigure 2. Properties of water, PCE and TCE as a function of temperature (from Heron et al. 2006).Other mechanisms include enhanced dissolution, hydrolysis, and aqueous phase oxidation.However, vaporization is dominant for most chlorinated solvents.In fractured rock systems, boiling of fluids in the fractures and the matrix leads to steam formation.The steam will sweep out of the rock towards locations with low pressure. Therefore,vacuum extraction is applied to each heater boring, creating a path for the generated vaporsout of the formation. By using each heater boring for extraction, it is ensured that the producedsteam can be extracted, and not migrate in unwanted directions. This principle is similarto the one developed for ISTD treatment of tight clay zones (LaChance et al 2006).SITE DESCRIPTION AND TCH DESIGNAt a site located in the southeastern part of the U.S., TCH was used to remediate a TCEDNAPL source zone that extended 90 ft below the ground surface (bgs). The bottom 15 feetof the treatment zone consisted of fractured gneiss (TerraTherm, 2007). In summary, the Sitewas underlain by 4 geologic units:• Fill: The fill was 25 feet thick, had a hydraulic conductivity of 1 x 10 -4 cm/sec, aporosity of 42 percent and a soil moisture content of 10.8 percent by weight.
• Saprolite: The saprolitic soil (severely weathered granitic gneiss) was 30 feetthick and had a hydraulic conductivity, porosity, and soil moisture content of 5 x10 -5 cm/sec, 40 percent, and 14 percent, respectively.• Partially Weathered Bedrock: Partially weathered rock (PWR) was present immediatelybeneath the saprolitic soil. It also had a hydraulic conductivity and porosityof 5 x 10 -5 cm/sec and 40 percent, respectively. This layer was 20 feet thick.• Fractured Bedrock: In general, the bedrock was assumed to be fractured, with ahydraulic conductivity of 1 x 10 -5 cm/sec and a fracture porosity of 0.5%. Thebedrock surface undulated with a typical depth to the bedrock surface of between75 and 80 ft. There was a possibility of a highly fractured zone beneath the targettreatment zone (TTZ) oriented north-to-south. The hydraulic conductivity of thehighly fractured zone was assumed to range between 1 x 10 -4 and 1 x 10 -3 cm/s andthe porosity was assumed to be 2%.The water table at the Site was at the bottom of the saprolitic soil at approximately 55 ft bgs,resulting in a total saturated thickness of approximately 25 feet of soil and partially weatheredbedrock overlying the fractured bedrock.The primary contaminant of concern (COC) present in the subsurface) at the Site that theISTD system was designed to treat was TCE. The TCE at the site was apparently released viaa sump/catch basin system associated with an aboveground TCE storage tank and a TCE reclamationunit (Tank Area). The amount of TCE released to the subsurface was unknown.The source area, i.e. the TTZ that the ISTD system was designed to treat was located adjacentto the southwest corner of the existing manufacturing building. The TTZ was selected in orderto encompass the highest soil concentrations and the most likely locations of TCE presentas DNAPL and included an area approximately 33 ft wide by 76 ft long (2,554 ft 2 ) with thelong axis oriented north-to-south.This alignment also coincided with the axis of the highest groundwater concentrations. Thedesign basis for the bottom of the TTZ was 87 feet below ground surface (bgs) to encompassvariations in the top of the bedrock and to ensure that all fill, saprolite, and weathered bedrockwithin the horizontal limits of the TTZ were treated. The heated interval extended to approximately90 ft bgs to ensure uniform heating of the bottom of the TTZ. The average depthto the top of bedrock observed based on the installation of the heater-only and heater-vacuumwells was approximately 79 ft. Thus all of the fill, saprolite, and weathered bedrock withinthe horizontal limits of the TTZ were treated. This amounted to approximately 8,230 cubicyards (cy) of soil and weathered bedrock.The primary remedial action objective for the ISTD installation was to remove TCE and otherCVOCs present from the unsaturated and saturated portions of the TTZ (i.e., above and belowthe water table within the TTZ) and to attain remedial standards. Although final remedialstandards were not established site-wide, the ISTD design was based on the achievement of60 μg/kg of TCE for soil in the unsaturated zone. Although similar treatment levels for soilbeneath the water table were feasible, the possibility that TCE present below the water table
outside of the TTZ could migrate back into the TTZ following treatment necessitated that nospecific remedial standard be set for the saturated portion of the TTZ. Instead, the ISTD systemwas designed to operate until the 60 μg/kg remedial standard for unsaturated soil was believedto have been achieved based on measurements of temperature and concentrations ofCVOCs in the well field vapor stream and interim soil and groundwater data. At that point,the ISTD system was to be shut down and the soil and groundwater present in the saturatedportion of the TTZ would be sampled and monitored to determine the level of cleanup achievedbelow the water table.Numerical simulations of the application of ISTD at the Site were performed prior to ISTDsystem design to provide a basis for development of the conceptual design. As a result of thenumerical simulations, a target treatment temperature of 212°F (100 o C) achieved in the interwellregions and the removal of a small fraction (i.e., 20%) of the water from the TTZ wasfound to be sufficient to achieve the remedial standards for TCE. Because the other CVOCspresent in the TTZ had similar physical and chemical properties (e.g., boiling points) as TCE,they were also found to be effectively removed from the TTZ by achieving 212°F (100 o C) inthe interwell regions. Thus, a target treatment temperature of 212°F (100 o C) was selected forthe project.A secondary remedial action objective of the ISTD installation at the Site was to minimize thepotential for contaminant mobilization during treatment. Given the information available forthe site at the time of the ISTD design, high concentrations of TCE and DNAPL were thoughtto be present in the subsurface. Thus, the ISTD system was designed to minimize the potentialfor contaminant mobilization outside of the TTZ both vertically and laterally. Specificaspects of the design that were added to minimize the potential for contaminant and DNAPLmobilization included:Hot FloorExtension of the heaters into the upper approximately 10 to 15 ft feet of the bedrock andboosting the power output of the bottom portion of the heaters in order to establish a “hotfloor.” The objective of the hot floor was to provide a barrier to vertical migration of the contaminantsas the contaminants would be volatized and extracted from the subsurface whencoming into proximity with the hot floor. 1 As described above, this resulted in the heatingand treatment of the upper 15 to 20 ft of the fractured bedrock.Establishment of Upward Vertical GradientsA low-flow extraction system was designed to slightly lower the groundwater table within theTTZ, thereby creating upward hydraulic gradients across the bottom of the TTZ. The creationof upward gradients across the bottom of the TTZ was designed to offset the downward forcesacting on the DNAPL and to provide an added level of security to ensure that the DNAPL didnot migrate downward.1 TerraTherm holds exclusive license to several patents for implementation of a hot floor during remediation to prevent verticalmobilization (U.S. Patent No. 5,997,214 and international patents granted and pending).
Perimeter Heater-Vacuum WellBecause there was a potential that COCs existed up to the edge of the TTZ, heater-vacuumwells were placed around the perimeter of the TTZ to ensure that vapors were pulled back towardsthe TTZ and not pushed outward. The well-field is shown in Figure 3. Figure 4 showsthe completed system, with fiberglass pipe manifold and a concrete vapor cover. A total of 24heater wells/borings were used, ten of which were also used for vapor extraction.Figure 3. Heater boring/well locations and thermocouple locations at the Site. Heater wells are open circles.Heater wells with applied vacuum and circles with a red center. Thermocouples are denoted by a red “T” withina circle.Figure 4. Completed ISTD system.
Electrical power (1,500 kW) for the ISTD remediation project was supplied from the existingplant building electrical service. TerraTherm’s electrical distribution panels and all downstreamequipment, which was configured for 480V, 3-phase, 4-wire service, was wired to thesecondary side of the transformer provided by the host facility.The vapor collection system at the Site consisted of a moisture knock-out pot and a vacuumblower to draw the heated vapors from the ground and convey them to the existing stack via afiberglass manifold piping system that was constructed by TerraTherm. The liquid condensatefrom the manifold piping system and the liquid extracted from the recovery wells waspiped to the existing groundwater treatment system at the host facility.RESULTSISTD operations ran continuously 24 hours per day, 7 days per week from the start of heatingon January 29, 2007 through the end of the heating period on June 20, 2007 and the finalISTD system shutdown on June 25, 2007.Thermocouples (TCs) were installed at 7 locations between the thermal wells throughout theISTD well field to monitor the soil heat up. These TCs were used to determine when the targettreatment temperature was attained within and at the top and bottom of the TTZ. Attainmentof the target treatment temperature was used to gauge when the ISTD treatment could bestopped. Each temperature monitoring location consisted of an array of thermocouples locatedat selected vertical intervals (e.g., 18, 40, 68, and 83 feet bgs).This vertical array of thermocouples enabled evaluation of the ISTD system treatment progressin the various geological layers found at the Site.250Average of Temp200150100IDMW9-83'bgsT1-HO1/HV4-83'bgsT2-HV4/HO1-83'bgsT3-HV7/HV4-83'bgsT5-HO3/HV6-83'bgsT6-HV6/HV10-83'bgsT7-HV10/HV6-83'bgs5001/29/2007 2/15/2007 3/8/2007 3/23/2007 4/14/2007 5/2/2007 6/2/2007DateFigure 5. Thermocouple temperature readings (°F) at the Site over the duration of operations at 83 feet bgs (representativeof temperatures in the bedrock).
Figure 5 shows the thermocouple temperature readings at the 83 feet bgs depth, which is inthe bedrock. Similar or higher temperatures were achieved at the shallower depths as well.The target temperature of the boiling point of water was generally achieved in the entiretreatment volume after approximately 100 days of heating. Since the mass removal continuedto be measureable and the power usage was lower than expected, the client chose to extendthe operational period by approximately 7 weeks after the initial heat-up. Figure 6 shows thecumulative mass removal curve along with the projected and actual energy usage.35000003000000Actual Energy UseProjected Energy UseVOC Mass Removed1400012000250000010000Energy Usage (kwh)20000001500000100000050000080 days110 days120 days8000600040002000Mass Removed (lbs)01/28/07 2/17/07 3/9/07 3/29/07 4/18/07 5/8/07 5/28/07 6/17/07DateAchieve100CThroughout TTZFigure 6. Mass removal and energy usage at the Site.ConfirmatorySoil Sampling0DesignBasisThe total mass of VOCs extracted from the subsurface in the vapor phase during the course ofthe project was approximately 11,590 pounds (5.8 tons). The total mass of VOCs extracted indissolved liquid phase was approximately 92 pounds. In addition to the VOCs extractedthrough volatilization and the dissolved phase, it is expected that some additional mass ofVOCs would have been eliminated in situ due to hydrolysis or other in-situ degradation processessuch as direct oxidation or pyrolysis.The total amount of energy used to reach the remedial goal after 110 days of heating was1,500,000 kWh. The amount of electrical energy expended per volume treated, was 182 kWhper cubic yard. The total operating time and amount of energy that was estimated to be requiredto heat up the TTZ and attain the remedial goal was 120 days and 2,600,000 kWh, respectively.Thus, the amount of energy actually used to heat up the TTZ was 60% less thanthe design. This indicates that subsurface heat losses to areas surrounding the TTZ were lower
than anticipated, and the applied energy was used efficiently to raise the temperature insidethe TTZ. This is great news for thermal remediation in fractured rock.After 110 days of heating, a total of 66 discrete soil samples were collected and of these, 10were duplicates. All of the duplicates showed agreement indicating very little sample variability.Soil and rock samples from within the treatment zone show a very thorough removalof contaminants. Measured starting concentrations of TCE were as high as 81,000,000 μg/kgand 1,100,000 μg/L in soil and water, respectively, and DNAPL was visually observed in soiland water samples. The post-remediation 95% Upper Confidence Limit (UCL) of the meanTCE soil concentration for the entire treatment zone, above and below the water table (basedon 56 discrete soil samples), was 17 μg/kg (TerraTherm, 2007). The post-treatment concentrationof TCE in groundwater samples from a monitoring well within the treatment zone thathad starting TCE concentrations at saturation levels (1,100,000 μg/L) was reduced to
LaChance, J., G. Heron and R. Baker. 2006. “Verification of an Improved Approach for Implementing In-SituThermal Desorption for the Remediation of Chlorinated Solvents.” Paper F-32. , in: Bruce M. Sass (ConferenceChair), Remediation of Chlorinated and Recalcitrant Compounds—2006. Proceedings of the Fifth InternationalConference on Remediation of Chlorinated and Recalcitrant Compounds (Monterey, CA; May 2006). ISBN 1-57477-157-4, published by Battelle Press, Columbus, OH, www.battelle.org/bookstore.TerraTherm, Inc. 2007. Remedial Action Completion Report. Implementing In-Situ Thermal Desorption (ISTD)Remediation. [Confidential SE US Site]. November.