~ Pergamon

PH: S0273-1223(98)OO259-5

War. Sci. Tech. Vol. 37, No.8, pp. 161-168, 1998.© 1998 fAwQ. Published by Elsevier Science Ltd

Printed in Greal Britain.0273-1223/98 S19'00 + 0'00

COST EFFICIENCY OF SUBSURFACEREMEDIATION USING SOIL VAPOREXTRACTION AND GROUNDWATEREXTRACTION

T. Hirata*, N. Egusa*, O. Nakasugi**, S. Ishizaka*** andM. Murakami***

* Department of Environmental Systems, Wakayama University, 930 Sakaedani,Wakayama City, Wakayamll 640, Japan** National Institute for Environmental Stullies. 16-20nogawa, Tsukuba City,lharaki305, Japan*** Kumamoto City Government, I-I Tedori-Honmachi, Kumamoto City,Kumamoto 860, Japan

ABSTRACT

The groundwater pollution due to volatile organochlorines like trichloroethylene and tetrachloroethylene hasbeen a great environmental issue in Japan. Thc nation wide survey revealed on the basis of up to fifty-ninethousand samples collected until 1995 that 1.5% for trichloroethylene and 2.3% for tetrachloroethylenecannot meet the drinking water standard. In order to repair subsurface pollution and to establish theintegrated procedure for cost-benelit remediation measure, physical remediation technologies of soil vaporextraction and groundwater extraction were applied to a study site contaminated with trichloroethylene. Theresults showed that the trichloroethylene amounts of 472 kg by soil vapor extraction and 1764 kg bygroundwater extraction were removed during three-year operation. In addition experience with bothtechnologies has demonstrated that the soil vapor extraction has been successful in removing I kg hr·1 oftrichloroethylene at the initial stage of remediation, which shows one order as high as the groundwaterextraction. However, the removal rate due to soil vapor extraction declines much earlier than groundwaterextraction, and consequently the removal rates of both technologies develop inversely with the progress ofremediation. Such remediation behavior of technologies raised the relative cost for soil vapor extraction 15times as high as groundwater extraction. © 1998 IAWQ. Puhlished hy Elsevier Science Ltd

KEYWORDS

Groundwater pollution; volatile organochlorine; trichloroethylene; remediation; soil vapor extraction;g.<'\.\\.~~,-....~~~i.).\1:,<:n.~ ....~"'-\,"...w-"."

INTRODUCTION

Large-sca{e groundwater po{(ution due to hazardous chemica{s was first discovered in a study by the JapanEnvironment Agency in {982 which traced the po{lution to the organoch{orines such as trich{oroethy{eneand tetrachloroethylene. The survey WU8 triggered by the discovery of groundwater pollution with volatileorganochlorines in developed nations throughout the world and the detection of volatile organochlorines inlapan, when general mai.n mpply water W'\l'. analyred. fQr tri.haIQmethane'i.. The s.ur'ley was ccmd.u.cted. tQ

161

162 T. HIRATA et al.

explore the groundwater pollution situation with regard to 18 substances in 1360 well waters sampled in 15cities across the country. The results showed that trichloroethylene and tetrachloroethylene were detected ata rate of one in three well waters, and three percent of groundwater samples overshot the drinking waterstandards of 0.03 mg I-I for trichloroethylene and 0.0 I mg I-I for tetrachloroethylene (Hirata et at., 1992).

Following the results of the nationwide survey, a large number of local authorities carried out their owninvestigations on the state of groundwater pollution with regard to three substances, the above two chemicalsplus I, I, I-trichloroethane. Until 1995 fifty-nine thousand groundwater samples were collected, and theresults showed the temporal trends of excess rates for the three substances over drinking' water standards asillustrated in Fig. I. Data are coming from a general situation survey to investigate the state of pollution bydividing the survey area in a mesh grid, and investigation sites are not duplicated at all. Since 1989, thestatistical incidence of excess rate for standards has reduced to be 0.3-0.6 % at present; however, this may betotally attributed to the fact that at the initial stage the nationwide surveys focused on areas with a high riskof pollution such as those surrounding industrial-commercial users of the chemicals and now are extendingto residential areas and the countryside with a lesser potential for pollution.

The groundwater pollution due to volatile organochlorines is confirmed to exist on a nationwide scale, andin addition the organochlorines like trichloroethylene are strongly resistant to biodegradation in subsurfaceenvironment (Vogel and McCarty, 1985); therefore, such pollutant removal from subsurface environment isinevitable for groundwater quality to meet drinking water standard. In this context various sorts of physical,chemical and biological remediation technologies have been developed and tested in polluted sites of Japan.However, each of them is highly costly. The success of the remediation operation totally depends on pickingup suitable technologies to reduce the volume of polluted soil, water and gas to be treated, corresponding tothe pollutant existing form in the subsurface environment. In order to establish the cost-benefit relationshipin remediation design a full scale remediation using soil vapor extraction and groundwater extraction hasbeen conducted in a volcano-ash soil region contaminated with trichloroethylene. The paper describes theremediation operation and cost-benefit of both technologies, based on the temporal variations ofconcentration and removal rate of the pollutant.

<> Trichloroethylene[] Tetrachloroethylene~ 1,1. 1-trichloroethane

6

,...." 5-~ 4t)...III 3"-

en2en

lU(J>< 1w

01983 1985 1987 1989 1991

Observation year1993 1995

Figure I. Temporal trends of excess rates for trichloroethylene. tetrachloroethylene and 1,1,1- trichloroethane overdrinking water standards.

STUDY SITE DESCRIPTION

Site investiiatjon

During the general monitoring of groundw,lter pollution, trichloroethylene above the drinking water standardwas detected from a domestic water supply well. At first surface soil gas reconnaissance was conducted toidentify pollutant source and highly residual area of trichloroethylene (Yoshioka et at., 1992), and resultedin revealing the trichloroethylene soil gas concentration of up to 330ppmv in a car park of a firm which

Cost eff'iciency of subsurface remediation 163

utilized the pollutant as solvent for electric components. At next stage according to the gaseous plume oftrichloroethylene in surface soil, borings were made at fourteen locations as shown in Fig 2. The boringinvestigation showed that the study site is covered by volcano ash over 60 m, overlying an impermeable claylayer.

........~7 1,..,

• firm

oL ,

L..!.-201:SOCI

H

4-I...I

s-..

Figure 2, Locations of borings and boreholes for remediation. The closed circles of B series are the locations forboring site and the closed triangles of K series for remediation dual extraction well. The lines in the figure denote

the altitude of the groundwater table from sea level.

Existing form of trichloroethylene

The soil concentration of trichloroethylene is displayed in Fig. 3, which is in a cross section along with B-7,B-5, B-6, B-II, B-1 to B-9 as in Fig. 2. As there are volcano ash deposits of up to 60m and seeing that thegeological profile shows little change, the existing form measurement provides a very clear picture of themigration behavior of organochlorines in soil. The maximum soil concentration reaches 138 mg kg -I at adepth of GL-46m below the ground surface of B-6 site. Similarly the maximum for groundwaterconcentration has been measured as 295 mg I-I at the same location. In addition it is also confirmed thattrichloroethylene was not detected from the deeper aquifer below the impermeable clay layer.

Taking an example on a contour line of 10 mg kg-I", the range of spread will not reach more than 45m evenwhen trichloroethylene percolates more than 40 m. In the case of a 100 mg kg-I contour, this range ofspread becomes narrower, extending only about 10m. It appears that trichloroethylene permeating into thesoil migrates almost straight down withollt any significant transverse spread. Such migration features oftrichloroethylene in the subsurface environment depend on physical/chemical properties, Le., volatileorganochlorine liquids are heavier in weight and lower in surface tension and viscosity than water, therefore,they are rather mobile in the vadose zone compared to water.

REMEDIATION OPERATION USING DUAL EXTRACTIONS

Since the highly residual part of trichloroethylene is existing in both sides of vadose and groundwater zones,the soil vapor extraction method was employed for vadose zone remediation with pumping up thegroundwater at the same time. In order to extract both soil gas and groundwater from the same remediationwell, four dual extraction wells of K-I to K4 were constructed around the hotspot for the soil gas flow reachto cover the highly polluted area (Hirata and Nakasugi 1993). The diameter of K-I is IOcm, and others are2Ocm, with each depth of 60m. The groundwater table is located at a depth of GL-42m, therefore, thescreened portion to take soil gas and groundwater was installed at the interval of 29m between GL-31 m andGL-60m.

164 T. HIRATA et al.

The extracted groundwater is subjected to an air-stripping treatment, and the pollutants contained in the airinjected into the stripping tower were adsorbed on activated carbon before releasing the spent air into theatmosphere. Similarly, the pollutants in extracted soil gas with suction inside the well exerted by highDower blower were also treated with activated carbon adsorption.

Time-varied changes of trichloroethylene in soil gas extracted

Prior to the remediation operation, the soil gas flow reach was examined using two remediation wells at theextraction pressure of 0.6atm on the basis of the theory by Johnson et ai. (1990). The test resulted in makingthe radius of influence by soil vapor extraction to be approximately 10m, which covers the highly residualparts of trichloroethylene by eighty percent.

Fig. 4 illustrates the time-varied changes of concentration of trichloroethylene contained in extracted soilgas. Since the K-9 well was constructed close to the hotspot, consequently the concentration takes themaximum among the four extraction wells. It appears that all of trichloroethylene concentration tend todecline linearly with operation time on the full-log scale. In addition the best fit lines for each soil gasconcentration were displayed. when the relationship of the concentration C and operation time t isformulated by the following equation.

(I)

The soil gas flow rate differs, depending on a slight change of geological situation, however, it tends to keepapproximately constant for each extraction well, i.e., 110, 190, 60 and 310 L min-I in order of the K-l wellto the K-4 well. This feature produces the temporal trend of trichloroethylene removal rate by soil vaporextraction similar to that of soil gas concentration with the operation time.

--==

--

~3~'"1

"i"l;;) ~_~·i~1:;_;====·I=-·==·i:·="::1 ":.....::'~i:'3==:===...::1=Surfece.. rea ...

~~........-_......--"III- "'0

Figurc 3. Soil conccntration or trichlorocthylene in case or rclalivcly deep subsurface pollution.

Time-varied changes of trichloroethylene in ~roundwater extracted

The trichloroethylene concentration of groundwater was confirmed from boring investigation to reach 294mg I-I in the neighborhood of K-2 well. In this context the groundwater of the trichloroethyleneconcentration being over 100 mg I-I was expected to come up at the initial stage of the remediation, andhence the total pumping rate was reduced to approximately 2 m3hr- 1 to fully reduce the groundwaterconcentration treated by air stripping method below the drinkable limit. The contribution of K-l and K-4

Cost eflieiency of subsurface remediation 165

wells to the total is 0.5 and 1.5 m3hr- l . During the operation time of 7060 to 10100 hours the groundwaterpumping rate wa~ intensified to 16.5 m3hr- l • Furthermore after the operation time of 10100 hours when thetrichloroethylene concentration of extracted groundwater decreased to the level of approximately 10 mg I-Ithe total pumping rate was raised to 28.9 m3Ilr- l , each contribution to which is 2.0, 03,3.4 and 23.2 m3hr- 1

in the order of the K-l well to the K-4 well.

10000

(PplIv)

......~en ..",-,-..f'0 1000II)

<:

~'"! 0 K-l C = 2. 84-1 05t -o.a79

'" y • 0.9852>0- 100 • K-2 c = 1. 45-tOSt-o. m;S y • O. 9801'"0 0 K-3 C • 22.3-t05t-1.%es0 y • 0.9712'fi • K-4 C • 1.16-105t-o. 998... y = 0.9815....

1010 100 tODD (hr) 10000

Operation t ime(t)

Figure 4. Time-varied changes of trichloroethylene concentration in soil gas extracted from four dual extractionwells.

Fig. 5 shows the time-varied changes of trichloroethylene concentration contained in the groundwaterextracted from two wells K-l and K-4. The values of a and b in the figure are constant, when thetrichloroethylene concentration depends on the operation time as in the form of Eq(I). The concentrationshave been steadily reducing, and in particular after the operation time of 10100 hours the pumping rateaccelerates the reduction speed of the trichloroethylene concentration. Then, the time at which thegroundwater concentration meel~ the drinking water limit of 0.03 mg I-I for trichloroethylene was estimated,using the exponent relationship for the K-4 well after 10 I00 hours noted as the equation (2) in Fig. 5. As aresult, the groundwater extraction to repair the pollution state should be continued for 31.3 years more.

COMPARISON OF COST BENEFIT OF BOTH TECHNOLOGIES

The remediation operation using two technologies resulted in removing the trichloroethylene amounts of 472kg by soil vapor extraction and 1764 kg by groundwater extraction during the period of 27700 hours (1154days). Fig 6 compares the time-varied changes of the trichloroethylene removal rate in the unit of kilogramsper hour: Both removal rates indicate the total rate of the four extraction wells. It appears that at the initialstage of the remediation, the soil vapor extraction operation has been successful in removing by suction I kghr l of trichloroethylene, which is an order of magnitude higher than the removal rate achieved withgroundwater extraction. Yet the removal rate due to soil vapor extraction does begin to decline much earlierthan in case of groundwater extraction, so that the removal rates of these two methods develop inverselywith the progress of remediation. In particular the removal behavior of pollutant totally depends on thechemical feature of trichloroethylene liquid being highly volatile and little soluble in water (Hunt and Sitar,1988) and the change of existing form of pollutant affects the effectiveness of applied remediationtechnologies with the operation time going on. The groundwater extraction requires so many years to

166 T. HIRATA et al.

cleanup the polluted aqueous phase, contmry to this, it is likely to eliminate much more pollutant withgroundwater extraction than other technologies. In this remediation operation, the additive effect of thegroundwater pumping rate after 10100 hours from the onset of the operation quickens the crossing time ofboth removal rates.

10000010000100 1000Operat ion time

10

1000(mg·!"1 )

l-II)... o 0<0

~ 0 0 0

5 100el-e) • ~c 08

.0\~ t)~'

do ~

c ~ \.e 10 ! ..... 0 ipc,) 0 K-lII) C' 4. 37Xl0Z%t-5.IT .. ~ \cII)

7 • 0.9492>-.s::. • K-4- (1) C • 87. 2t -o· 2&1II)

~e

7 • 0.6535l-e (2) C • 3.55 x 1CSt-I. 301.s::.c,) 7 • 0.5500I-~

(hr)0.1

Figure 5. Time-varied changes of trichloroethylene concentralion in groundwater extracted from Iwo dualextraction wells or K-I and K4.

The remediation operation is highly costly, so that it is inevitable from the cost-benefit point of view toestablish integrated procedure of remediation measure including a suitable survey method and pickingeffective technology up. Taking an example on the data presented in Fig. 6, the running costs required forthe groundwater extraction operation in the first month of the remediation program have been considered asunity and the monthly changes of the normalized relative costs for groundwater and soil vapor extractionsare plotted in Fig. 7. The running costs comprise electricity, water supply, exchange of exhausted activatedcarbon, chemical analyses and control of the operation systems. With the passage of time, the removal ratesfor both of these operations are declining, and the relative costs for both operations are increasing until theend of the eighth month. Conversely the groundwater pumping rate increased substantially from thethirteenth month and after this the relative cost for soil vapor extraction has still continued to rise and finallymounts to the value of 16, although that for groundwater extraction has fallen below unity. As a result, theremediation measures at this pollution site have led to the discontinuation of the soil vapor extractionoperation due to the declining removal rate and the rising relative cost, while the groundwater extractionoperation is currently continued. The result implies that from cost-beneficial point of view the application ofa single remediation technology to polluted site is limited to reach the final goal.

CONCLUSIONS

The success of remediation operation is attributed to determining the detailed pollutant existing form in soiland groundwater environments. In other words, it is required to delineate the area with a pollution risk onthe basis of groundwater quality surveys and soil vapor monitoring and to enforce detailed surveys in order

Cost elTiciency of subsurface remediation 167

to identify the pollutant intrusion locations to ground in the area where the pollutant has been discovered.With respect to volatile pollutants, at locations with high concentration of soil vapors, the pollutant intrusionis presumed to have occurred so that boring survey will be carried out to explore the vertical profile ofpollution and to complete three-dimensional structure of pollution.

41 Groundwater extraction(1) R • O. 574t·o.416

V • 0.8453(2) R • 4. 30x 10't'Z, 58

V • 0.8628

10 O. 1>

~.,c.,>--5., O. 01~o

.r.<J

0.00110

'Q.

100

o Soi I venting R· 24.5t -t.8O%V • 0.9784

•

•

ChI')

1000 10000Operation timeCt)

100000

Figure 6. Comparison of trichloroethylene removal rates due to soil vapor and groundwater extractions., Thenumerals a and b in the figure denote the conslants, when the removal rate R is fonnulated with the operation time t

by R = atb.

-<>-Groundwater extraction--Soi I vapor extraction

1816

+'" 14III

120uG) 10> 8+'"co 6G)

c::: 420

0 2 4 6 8 10 12 14Operation time (month)

16 18

Figure 7. Monthly changes of relativc cosl efliciency for soil vapor and groundwater extractions based on therunning costs al the first month needed for groundwater extraction.

After the existing forms have been revealed by a series of pollution surveys, the next step is to determine theremediation technology to be applied in response. The technique adopted depends on the scale and depth ofpollution. There is no effective remediation technology for every form of pollution. It is also difficult toapply a single remediation technique throughout the operation from the initial stage to the final stage of theremediation, as recognized in the comparison of relative costs for the operations of soil vapor and

168 T. HIRATA et al.

groundwater extractions. In accordance with the progress of remediation, it is essential to keep a flexibleapproach so as to change over to a more cost-efficient or low-cost technology as the measure proceeds.

REFERENCES

Hirata. T.. Nakasugi. 0 .. Yoshioka. M. and Sumi. K. (1992). Groundwater pollution by volatile organochlorines in Japan andrelated phenomena in the subsurface environmcnt. War. Sci. Tech.. 25( II). 9-16.

Hirata. T. and Nakasugi. O. (1993). Remedilll operation for subsurface pollution due to volatile organochlorine using soilventilation and groundwater extraction. In: COlltaminared Soil '93. F. Arendt et al. (ed.). Kluwer Academic Publishers.pp.1019-1028.

Hunt. 1. R. and Sitar. N. (1988). Nonllqueous phllse liquid transport and cleanup. I. Analysis of mechanisms. Wat. Resour. Res..24(8). 1247-1258.

Johnson. P. c.. Kemblowsld. M. W. and Colthart. 1. D. (1990). QUllntitative analysis for the cleanup of hydrocarbon-contaminatedsoils by in-situ soil venting. Groulld Warer. 28(3). 413-429.

Vogel, T. M. and McCarty. P. L. (1985). Biotransformation of tetrachlorocthylene to trichloroethylene dichloroethylene. vinylchloride. and carbon dioxide undcr methanogenic conditions. Appl. Ellvimll. Microbiol.• 49. 1080-1083.

Yoshioka. M., Yamasaki. M.. Okuno. T.• Hirata, T. and Nakasugi. O. (1992). Soil gas monitoring for survey on groundwaterpollution due to volatile halogenated hydrocarbons. J. Japan. Society Waf. Environ.. 15(10). 719-725 (in Japanese).