1

Soil Vapor WorkshopSoil Vapor WorkshopMarch 2007March 2007

Field & Fixed Laboratory AnalysisField & Fixed Laboratory Analysisfor VI Investigationsfor VI Investigations

Blayne HartmanH&P Mobile Geochemistry

Carlsbad, CA760-804-9678

bhartman@handpmg.comwww.handpmg.com

This presentation is an excerpt from the vapor intrusion training that Dr. Hartman has been presenting to Federal & State regulatory agencies, DOD facilities, consulting groups, and stakeholders around the country.

Lecture notes are at the bottom of each slide so that if played out as a hard-copy, the presentation can be a useful reference document.

2

Get Enough DataGet Enough Data• Soil Gas Not Homogeneous• Spatial & Vertical Variations Exist • Don’t Chase 1 pt Anomalies• Get Enough Data Near/Around/Under• On-site Analysis Enables Real-Time

Decisions to Select Additional Sampling Locations

Soil gas, like soil, is not homogenous in most cases. So you need enough data to give decent coverage near, around, or under the receptor. Simpler collection systems with small volumes are advantageous as there is less to go wrong and enable higher production per day (20+ samples per day). Less expensive analytical methods (8021, 8260) enable more analyses for reasonable cost. Real-time data can be extremely helpful to track soil gas contamination laterally and vertically.

3

Common Soil Gas Analyses Common Soil Gas Analyses • VOCs

– Soil & Water Methods: 8021, 8260– Air Methods: TO-14, TO-15, TO-17

• Hydrocarbons– 8015 m, TO-3

• Oxygen, Carbon Dioxide– ASTM 1945-96

• SVOCs: TO-4, TO-10, TO-13 (8270?)

This slide gives a summary of the most common analytical methods used for soil gas samples. More discussion on these methods follows.

4

Portable Field Analyzers Portable Field Analyzers • VOCs

– Hand-held PIDs - 1 to 10 ppmv– Mini-Rae (ppbv) – 50 to 100 ppbv

• Hydrocarbons– Portable FIDs (Foxboro, Photovac)

• Oxygen, Carbon Dioxide, Methane– LandTech GEM 2000

• Methane: Numerous

A variety of portable field analyzers exist for analyzing soil gas samples. For vapor intrusion applications, required detection levels of VOCs are lower than the field analyzers can reach, but they still are useful for screening sites or for sites with high concentrations. These meters give a total concentration only, so they are limited at sites with more than one compound.

Portable hydrocarbon detectors have higher detection limits (~5 ppmv) and do not give compound resolution.

Oxygen, carbon dioxide, and methane can be reliably measured with field meters to required detection levels.

5

TVA 1000 PID Readings vs TAGA Unit

y = 0.8875x + 1.189R2 = 0.9745

0

20

40

60

80

100

120

0 20 40 60 80 100 120

TAGA value (ppmv)

TVA

1000

PID

read

ing

(ppm

v)

This plot shows a comparison of data measured with a portable PID (Foxboro TVA 1000) to data measured with an on-site lab (TAGA) at a test site in Indianapolis in August 2006. The agreement is quite good for values exceeding 4 ppmv.

6

Soil Gas VOC Analysis Soil Gas VOC Analysis (TO(TO--14/15/17 or 8260 or 8021)14/15/17 or 8260 or 8021)

• All Methods Give Reliable Results

• Detection Level Discriminator– TO Methods: <1 to 10 ug/m3 $200-$300– 8260 SIM: 10 to 50 ug/ m3 $100-$150– 8021: 50 ug/m3 $75 to $100

• On-Site Analysis– Extremely Helpful for VI – Minimizes False Positives

A variety of analytical methods are available to measure soil gas samples. No federal guidance document exists specifying any one. Methods 8021 and 8260 are soil & water methods but give accurate results for soil gas samples at detection levels above 10 ug/m3. The toxic organic methods (TO) are designed for ambient air samples, so they give accurate results for soil gas samples at much lower detection levels. The TO methods require extensive hardware and are far more expensive. The criteria for selection should be which method(s) reach the required detection limits.

On-site data are extremely useful to ensure that the samples do not have tracer/leak levels above acceptable levels, provide real-time data for decision making, and to validate detections seen in the off-site data. If measured values are high, then the on-site methods (8021, 8260) are more appropriate to use than the ultra-sensitive TO methods. If on-site values are low or below detection, then the samples can be measured off-site by the TO methods.

7

Mobile Laboratory CapabilitiesMobile Laboratory Capabilities• VOCs

– 8021, 8260, Air Methods: TO-14, TO-15• Hydrocarbons

– 8015 m, TO-3• Oxygen, Carbon Dioxide

– ASTM 1945-96, Method 3• SVOCs: 8270

Same QA/QC Specs as Fixed Labs

Mobile laboratories are now capable of performing many sophisticated analyses previously only possible in fixed labs. This is due primarily to advances in analytical instrument technology. Simple GC methods as well as GC/MS methods, including the TO-methods, are now possible to do in mobile labs. The QA/QC criteria are no different than those in fixed base laboratories.

8

OnOn--site 8021 vs.offsite 8021 vs.off--site TOsite TO--15 15

Summa Canister - TO-15

1 10 100 1000

Tedl

ar B

ag -

On-

Site

GC

1

10

100

1000

This slide shows a comparison of on-site analysis of TCE by 8021 out of a tedlar vs. off-site analysis by TO-15 out of a Summa canister collected by EPA-ORD at a test site in CT. Correlation is excellent down to values as low as 2 ppbv.

Slide courtesy of Dr. Dominic DiGuilio, EPA-ORD

9

OnOn--site 8021 site 8021 vs vs TAGATAGATAGA vs. 8021 TCE concentrations

0

10000

20000

30000

40000

50000

60000

70000

0 10000 20000 30000 40000 50000 60000 70000 80000 90000

TAGA TCE (ppbv)

8021

TC

E (p

pbv)

This plot shows a comparison of results from on-site 8021 to data measured with an on-site lab (TAGA) at a test site in Indianapolis in August 2006. The agreement is quite good for values down to 10 ppbv.

10

OnOn--site 8260 site 8260 vs vs offoff--site TOsite TO--1515

This plot shows a comparison of data measured by 8260 from 60 cc syringe samples to data measured with an on-site lab (TAGA) from 1 liter tedlar bag samples at a site in Los Angeles. The agreement is quite good for values exceeding 100 ug/m3..

11

High SG Concentrations High SG Concentrations Create HeadachesCreate Headaches

• Typical Soil Gas Concentrations– Benzene near gasoline soil: >100,000 ug/m3– TPH vapor: >1,000,000 ug/m3– PCE under dry cleaner: >100,000 ug/m3

• TO-15 Maximum Conc: 2,000 ug/m3– Must do large dilutions, DL goes up– False positives from hot samples

• Canister & Hardware & Instrument Blanks

Typical soil gas concentrations at leaky UST, dry cleaner, and industrial solvent sites are in the 100,000s to 1,000,000 of ug/m3. But, for 1 in 1 million risk , the risk-based screening levels are less than 10 to 100 ug/m3. This large concentration range creates a number of analytical headaches.

The TO-methods and hardware (canisters, flow chokes) are not designed for such high concentrations. System carryover, large dilutions, and contaminated canisters increase the potential for false positives, raises reporting levels, and gives air labs logistical fits which limits the utility of these methods.

The 8260 and 8021 methods can’t get lower than 10 to 100 ug/m3 so they may not reach required DLs.

In practice, a combination of these methods is the best approach. If expected values are high, then the 8021 & 8260 are advantageous to use than the ultra-sensitive TO methods. If expected values are low, then the TO methods offer advantages.

12

Here is an example of what happens to reporting limits with highconcentration soil gas samples. The value for PCE is 140 ug/L (140,000 ug/m3). The TO-15 method upper limit was 100 times lower, so the sample had to be diluted 100 times. Consequently, the reporting limit for all compounds increased to levels higher than could have been reported by Method 8260. This situation is common at UST sites, dry cleaners, and commercial facilities using solvents.

13

Not All TONot All TO--15s Are Alike15s Are Alike

• Standard Method QA/QC Poor – Does Not Meet Many States Requirements – Can use standard for a year!!– No second source standard – No surrogates– Wider calibration acceptance windows

• Beware the “Wal-Mart TO-15”– Only use labs that have upgraded method– Only use lab that has a certification

The TO-15 analytical method has been advertised as the “Gold Standard”, but actually, the QA/QC is very poor and does not match the requirements of many State agencies. Further, it may have difficulty meeting the legal challenge. Incredibly, most State regulatory personnel don’t realize this. Some States, like NJ, have published a more exacting method than the standard method.

The higher-quality labs have upgraded the method to meet more exacting requirements required by the EPA SW-846 methods or any specific State regulatory requirements, such as second source standards & surrogates.

Beware the “Wal-Mart TO-15”. To ensure that you are getting a quality analysis, only use labs that can show they have upgraded the method QA/QC and have a certification from some NELAC or a State agency.

14

OnOn--Site/OffSite/Off--Site AnalysisSite Analysis

• Allows Measure of Leak Compound• Allows Assessment of Proper Purge Vol • Real-Time Results to Guide Program• Inconsistent Data can be Recollected• Hardware Can Be Reduced/Eliminated• Allows Optimal Method to Be Used

– 8260 if > 1 ug/L, TO-15 if < 1 ug/L

On-site data are extremely useful for a number of reasons: to ensure that the samples do not have tracer/leak levels above acceptable levels, to determine the proper purge volume, to provide real-time data for decision making, and to validate detections seen in the off-site data. Inconsistent results can be identified and corrected while still o-site. If measured values are high, then the on-site methods (8021, 8260) are more appropriate to use than the ultra-sensitive TO methods and much less costly. If on-site values are low or below detection, then the samples can be collected & measured off-site by the TO methods.

15

This is the typical profile for hydrocarbon VOCs, CO2, and oxygen in the soil gas with depth where bioattenuation is active. But how do you know at what depths this is occurring so how do you know where to sample? The depths are site specific, so the only way to know is after you collect the data. Real-time data allows you to identify this zone and collect samples at the proper depths to prove bioattenuation is active. If samples are analyzed off-site, you won’t know where to collect the samples when you are out in the field, so you must either go back for a second round of sampling or collect additional samples at shallower depths during the first round.

Theoretical Bio ProfileTheoretical Bio Profile

soil surface

O2

CO2

VOCscleansoil

petroleumproduct

increasingdepth

VOCs

O2

flux

16

SUB-SLAB VAPOR PROBE LOCATIONS AND RESULTS

REVIEWED BY:LC SCALE: Not to Scale

Approximate Sub-Slab Soil Vapor Probe Location (September 2006)

TVP-22

TVP-21

TVP-20

TVP-25TVP-24

TVP-23

TVP-19 TVP-18

TVP-16

TVP-17

TVP-13TVP-14

TVP-15

TVP-12

TVP-7

TVP-9

Sam ple # Benz SS RBSLTVP-12 25 840TVP-7 13 840TVP-9 21 840

Sam ple # Benz SS RBSLTVP-13 20 84000TVP-14 11 84000TVP-15 19 84000

Sam ple # Benz SS RBSLTVP-16 44 840TVP-17 <25 840TVP-18 40 840TVP-19 26 840

Sam ple # Benz SS RBSLTVP-25 20 840TVP-24 13 840TVP-23 48 840TVP-22 22 840TVP-21 15 840TVP-20 10 840

Sub-Slab Soil Gas Screening LevelsResidential above garage – 840 (1 e-6 risk)Commercial on slab – 840 (1 e-5 risk)Commercial above garage – 84,000 (1 e-5 risk)

Note – all levels are in ug/m3

This is a real job site where three different structures were assessed. Without real-time analysis, when would you have known to stop or to add more locations?

17

Coming Attractions:Coming Attractions:OnOn--site TOsite TO--15 Scan/SIM15 Scan/SIM• Simultaneous Scan/SIM mode enables

<10 ug/m3 for All VOCs & < 2 ug/m3 for subset of compounds.

• Only 2cc of Sample. Eliminates Hardware• Real-time Analysis in Structures: Control!• Already in CA

New equipment allows on-site TO-15 analyses. New GC/MS equipment enables simultaneous Scan/SIM mode meaning you can measure for all VOCs (>60 compounds) at DLs < 10 ug/m3 while simultaneously measuring for a subset of compounds at lower detection levels (<2 ug/m3). Only 2 cc of sample are required for analysis, so much of the sampling hardware can be eliminated, reducing chances of false positives.

This capability enables real-time analysis in structures. This might eliminate need for unsupervised time-integrated sampling.

18

Coming Attractions:Coming Attractions:Automated AnalyzersAutomated Analyzers

• GC and GC/MS (TO-15)• Can Reach Ultra-Low Levels (1-10 ug/m3)

for Subset of Compounds• Can Analyze 3 to 4 Times per Hour• Multiple Locations• Real-Time Feedback

Automated continuous analyzers exist that run unattended enabling analysis around the clock. They can sample from multiple rooms or probes and can send data over the internet in real time. Both gas chromatographs (GC) running 8021 and mass spectrometer (GC/MS) instruments running TO-15 are currently available.

19

Automated Analyzer DataAutomated Analyzer DataHUNTINGTON BEACH SITE - SOIL GAS

0

5

10

15

20

25

30

35

6:18:0

0 PM

7:09:0

0 PM

8:00:0

0 PM

8:50:0

0 PM

9:41:0

0 PM

10:32

:00 P

M

11:23

:00 P

M

12:13

:00 A

M

1:04:0

0 AM

1:55:0

0 AM

2:45:0

0 AM

3:36:0

0 AM

4:27:0

0 AM

5:18:0

0 AM

6:08:0

0 AM

6:59:0

0 AM

7:50:0

0 AM

8:41:0

0 AM

9:31:0

0 AM

10:22

:00 A

M

11:13

:00 A

M

12:03

:00 P

M

12:54

:00 P

M

1:45:0

0 PM

2:36:0

0 PM

3:26:0

0 PM

4:17:0

0 PM

5:08:0

0 PM

5:59:0

0 PM

6:23:0

0 PM

Time

Valu

e (%

) CH4CO2O2

This is an example of continuous monitoring of soil gas data, in this case methane, oxygen, and carbon dioxide in a vapor well. Similar data collected inside a structure could be useful in differentiating between vapor intrusion and ambient/background scatter. Or it can be used to demonstrate the repeatability of shallow soil gas data.

20

Location #A-4ft-BGS

0

50

100

150

200

0 10 20 30 40 50 60 70 80Time [d]

0

5

10

15

20

25

[v/v

%]

TPH [mg/L]CO2 [% v/v]O2 [% v/v]

Location #9-Sub-Slab- 0.5 ft BGS

0

50

100

150

200

0 10 20 30 40 50 60 70 80Time [d]

0

5

10

15

20

25

[v/v

%]

TPH [mg/L]CO2 [% v/v]O2 [% v/v]

Rea

lR

eal -- T

ime

GC

Sam

plin

gTi

me

GC

Sam

plin

gSource zone depth concentrations appear to be relatively stable with time

Sub-slab/near-crack variability with time

This is an example of continuous monitoring of TPH, oxygen, and carbon dioxide in two soil vapor probes using an automated analyzer for a period of 70 days from late September to late-November at a test site in WY. Soil gas variations at a depth of 4’ bgs at an exterior location showed little variation. Sub-slab soil gas concentrations immediately under a crack in the slab showed much larger variations.

Slide courtesy of Chevron/Texaco. Prepared by Paul Johnson, Arizona State University.

21

EPA NERL SV Test ProgramEPA NERL SV Test Program• Study Temporal Variations & Rain effects• Automated GC Running 8021• Thirteen Soil Vapor Probes Every Hour • Calibration Std & Blank Every Hour• Meterological Parameters • Running Since Last Friday• Real Time Feedback

Let’s Take A Look At the Instrument!

The EPA’s Las Vegas Research Lab has funded a study of temporal variations and effects due to rain on soil gas concentrations. An automated gas chromatograph is running 8021 analyses by itself on an array of thirteen soil vapor probes every hour. Meterological parameters such as wind speed, barometric pressure, temperature, and differential pressure are also being measured. The program will go on for at least a week and possibly longer.

22

EPA SV 3-07 22

Vandenberg AFB Study Site. Mobile lab in center. San dune environment. Ground water contamination of TCE at ~18 to 20 feet bgs.

23

EPA SV 3-07 23

Mobile lab showing 1/8” tubing from the soil gas probes running through electrical conduit into the van window.

24

EPA SV 3-07 24

Automated instrument with 16-port stream selector valve.

25

EPA SV 3-07 25

View into the lab showing laptop and printer. Laptop communicated with home office using broadband wireless cell service and PC Anywhere. Enables real-time transmission of data to office or home.

26

EPA SV 3-07 26

Weather station with satellite link installed by Tetratech EMI personnel to measure meterological parameters and vadose zone pressure.

27

VFB Soil Gas DataVFB Soil Gas DataProbe A3 (TCE)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

13:50

:09

17:50

:12

21:50

:15

1:50:1

9

5:50:2

2

9:50:2

5

13:50

:29

17:50

:32

21:50

:35

1:50:3

9

5:50:4

2

9:50:4

5

13:50

:49

17:50

:52

21:50

:56

1:50:5

9

5:51:0

5

9:51:1

0

13:51

:21

17:51

:27

21:51

:35

1:51:4

3

5:51:5

2

9:52:1

9

13:53

:39

17:53

:42

21:53

:45

1:53:4

9

5:53:5

2

Time (3/16/07 to 3/21/07)

Con

cent

ratio

n (p

pmv)

P

PP

This is a plot of the data recently collected by an automated instrument at the Vandenberg AFB site from three probes at the same location but at different depth (3’, 8, & 17’ bgs). This plot consists of almost 100 points per probe collected once per hour over a 4 day period. The soil gasconcentrations varied by less than 5% over these four days even for probes only 3 feet below the surface.

The study is still ongoing.

28

VFB Soil Gas DataVFB Soil Gas DataProbe A3-3' Port 9 (TCE)

0.15

0.155

0.16

0.165

0.17

0.175

0.18

0.185

0.19

0.195

0.2

13:50

:09

17:50

:12

21:50:1

5

1:50:1

9

5:50:2

2

9:50:25

13:50

:29

17:50:3

2

21:50

:35

1:50:3

9

5:50:42

9:50:4

5

13:50

:49

17:50:5

2

21:50

:56

1:50:5

9

5:51:05

9:51:1

0

13:51:2

1

17:51

:27

21:51

:35

1:51:43

5:51:5

2

9:52:1

9

13:53

:39

17:53

:42

21:53

:45

1:53:4

9

5:53:5

2

Time (3/16/07 to 3/21/07)

Con

cent

ratio

n (p

pmv)

Pro

This is an enlargement of the plot for probe A3-3’. This plot consists of almost 100 analyses collected once per hour over a 4 day period. The soil gas concentrations varied by less than 5% over these four days at this probe only 3 feet below the surface. Despite the minor variation, there appears to be periods of slightly decreasing and increasing concentrations. Is it real? . The study is still ongoing.

29

VFB Calibration DataVFB Calibration DataCalibration Gas - Port 1 (TCE)

0.5

0.7

0.9

1.1

1.3

1.5

13:18:08

17:18

:12

21:18

:15

1:18:1

8

5:18:22

9:18:25

13:18

:28

17:18

:32

21:18:3

5

1:18:3

8

5:18:41

9:18:4

5

13:18

:48

17:18

:51

21:18

:56

1:18:5

9

5:19:0

4

9:19:0

9

13:19:2

0

17:19

:27

21:19

:34

1:19:4

2

5:19:5

1

9:20:17

13:21

:38

17:21

:41

21:21

:45

1:21:4

8

5:21:5

1

Time (3/16/07 to 3/21/07)

This is a plot of the instrument calibration over the same 4 day time period. The standard was measured every hour and the measured concentrations varied by less than 5% over these four days.

The study is still ongoing.

30

VFB Soil Gas DataVFB Soil Gas DataProbe A3 (3/16/07 to 3/20/07)

0.725

0.73

0.735

0.74

0.745

0.75

0.755

0.76

0.765

0.77

0.775

0.78

1.44 1.46 1.48 1.5 1.52 1.54 1.56 1.58 1.6 1.62

Probe A3-17' Port 5

Prob

e A

3-8'

Por

t 10

TCE

This is a cross-plot of the results probe A3 at 8’ bgs vs. probe A3 at 17’ bgs. The good correlation suggests that the processes influencing the soil gas concentrations are consistent for both depths.

The study is still ongoing.

31

VFB Soil Gas DataVFB Soil Gas DataProbe A3 (3/16/07 to 3/20/07)

0.725

0.73

0.735

0.74

0.745

0.75

0.755

0.76

0.765

0.77

0.775

0.78

0.176 0.178 0.18 0.182 0.184 0.186 0.188 0.19 0.192 0.194 0.196 0.198

Probe A3-3' Port 9

Prob

e A

3-8'

Por

t 10

TCE

This is a cross-plot of the results probe A3 at 3’ bgs vs. probe A3 at 8’ bgs. The poor correlation suggests that the processes influencing the soil gas concentrations are different at the 3 foot depth compared to the 8 foot and 17 foot depth. Infiltration of atmospheric air? Even if so, the concentration variations are insignificant (<5%) and below analytical precision.

The study is still ongoing.

32

The Most Common GoofThe Most Common GoofWith Air & Soil Gas SamplesWith Air & Soil Gas Samples

1 ug/L Benzene equals: a) 1 ppbvb) 1 ppmvc) 330 ppbvd) None of the Above

The most common goof with air and soil gas samples is confusion over units. Let’s see how you do.

33

VI Analytical Documents VI Analytical Documents

• Overview of SV Methods (www.handpmg.com)– LustLine Part 1 - Active Soil Gas Method, 2002– LustLine Part 3 - FAQs October, 2004– LustLine Part 4 – Soil Gas Updates, Sept 2006

• Regulatory Guidance/Documents– ITRC VI Guidance - toolkit (www.itrcweb.org)– LA-RWQCB Soil Gas Guidelines – San Diego County Protocols

A summary of existing documents on soil gas methods can be found at these locations:

34

VI Analytical Documents VI Analytical Documents • API Gas Document (www.api.org/bulletins)• Comparison of Methods 8260 vs TO-15:

– Picker, H&P (www.handpmg.com)– Hayes, Air Toxics (www.airtoxics.com)

More documents.

35

EPA SV 3-07 35

Blayne Hartman, Ph.D.2470 Impala Dr.

Carlsbad, CA 92010(760) 804-9678

www.handpmg.com