This presentation is an excerpt of the vapor intrusion training that Dr. Hartman has been presenting to Federal & State regulatory agencies, DOD facilities, consulting groups, and stakeholders around the country. As of January 2010, this training has been given to over 30 State Regulatory agencies, including ASTSWMO and the State Coalition of Dry Cleaners. Training has also been given to many PRPs such as the major oil companies, Armed Services, & EPRI.

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.

1

The most important ingredient for cost effective and efficient VI investigations is the experience of the person/firm doing the collection. Is the collection being done by a firm that has prior experience? Is it a routine part of their services or an occasional part? Do they put experienced people in the field who can think or junior staff who aren’t well versed? This applies to the consultant and their subcontractors.

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 verticallycontamination laterally and vertically.

Legal considerations often dictate what additional work needs to be done at what standards.

All of these issues affect the investigation progress.

2

g p g

The two biggest goofs:

3

Here is a summary of existing Federal & State guidances as of September 2010.

4

Each investigatory approach has pros and cons that must be considered before choosing the one to use at a site.

5

The most common model currently being used in the Johnson & Ettinger (J-E) model. The EPA has written a number of spreadsheets that are based upon the model parameters and allow the use of groundwater, or soil, or soil gas data. The spreadsheets were updated in 2003 and are available from the EPA website referenced previously. Many agencies have constrained many of the model parameters so they can’t be “abused”.

The big issue at this time is whether you can close the pathway from these spreadsheets and exterior data only.

6

Some variables are easy to change in the spreadsheets, while some others are not.

7

8

There are issues that need to be considered when sampling indoor air and when interpreting the data. Sampling issues include the hardware, time period for collection, and things as simple as the type of marker used to label the samples.

9

Some of the issues that need to be considered when installing probes include:

Tubing Type: Small diameter tubing offers advantages over large PVC pipe. Flexible tubing tends to leak.

Probe tip: Metal tips may have blanks due to the cutting processProbe tip: Metal tips may have blanks due to the cutting process.

Equilibration time: How long to wait, especially if air knives are used to clear holes or larger drill rigs are used?.

Equipment blanks: need to collect blank through the collection system. TripEquipment blanks: need to collect blank through the collection system. Trip blanks not enough.

10

Lower detection levels requires more careful protocols. Important sampling considerations include sample volume, container type, flow rate, and leak testing to ensure valid samples are collected. Smaller volumes require less complicated sampling systems and minimize the chances for leakage from the surface and desorption off soil. Recent studies have shown no difference in soil gas values regardless of whether small (0.5 L) or large (100 L) volumes are collected.Sample containers must be inert, clean, and handled properly (no cooling or heat). Canisters have longer holding times, but have the potential for blanks (carry-over from previous samples), cost more, and can be trickier to fill. Tedlar bags are good for ~2 days, are less expensive, and suitable for concentrations of 1 ppbv or higher.

Sample flow rate is of concern to many agencies, but recent data are showing it not to be a factor.

Tracer/leak compounds are generally required to ensure sample integrity because small leaks can create significant effects at such low concentrations. The larger the volume extracted and the more complicated

11

g pthe sampling system, the greater the potential for leaks.

Smaller canisters fill more quickly and are easier to handle. Tedlar bags are simple to use and not prone to as many problems as canisters.

12

There are some differences in soil gas sampling for petroleum hydrocarbon VOCs than for chlorinated solvents. If samples at deeper depths exceed allowable values, shallower samples may need to be collected to document the effect of bioattenuation. Oxygen data should always be collected to document presence of the aerobic zone. Soil phase data may be needed to document the presence of a clean soil layer.

13

These are things you want to do/allow to practically and cost effectively assess this risk pathway.

14

Upcoming vapor intrusion training in 2010

15

A real-live case history of how not to do a vapor intrusion assessment.

16

Note the location of the 25 foot deep sample that is being used to evaluate the vapor risk at this site. There are a number of other soil gas sample probes adjacent to this building; none of these closer sample probes had hydrocarbon levels above action levels demonstrating that the VI pathway is not complete at this site. Why was this “common sense” approach not used at this site? Why was indoor air sampling necessary to evaluate if there was risk from upward migrating vapors at this site?

17

Note the units for the samples in this table: ug/L Note the samples at 5’ & 15’ are much lower than the 25’ sample.

Also, note the large difference in sample results as a function of purge volume. This demonstrates that while the concentration can be high the volume is low otherwise you would not observe this large of a difference in results as a function of the low amount of gas that is being purged to collect this sample. Several more purge volumes might have removed all the hydrocarbons; we have observed this at a number of sites.

18

The consultant used the DTSC guidance document, not the local oversight agencies’ guidance, and concluded that there is not a risk to indoor air at a nearby building. But the consultant used the non-cancer allowed levels for benzene, not the cancer risk levels. So they reached the wrong conclusion.

19

The preliminary screen used a default attenuation factor of 0.001 and shows in this table that the site passes.

But wait a minute, the units are different from Table 1.

If the measured result was really 1,200 ug/L then the correct number to use in this table would be: 1 200 000 ug/m3table would be: 1,200,000 ug/m3.

Which units are correct?

20

The consultant ignores his contradictory text re the results of the preliminary screening and the site-specific screening.

Site specific data collected from samples at shallower depths near the building documented that the pathway was not complete and there was no risk from upward migrating hydrocarbon vapors. So why do indoor air sampling?

Especially for benzene in an urban environment?

21

Although concluding in the text that it was not necessary to go to the next step, the consultant did so anyway. The consultant used the J-E model to determine the risk from the soil gas data. First, they used the wrong version of the spreadsheet. This time they reached the opposite conclusion. This is odd since site-specific screening is generally less conservative than preliminary screening. What happened? They now made the comparison based upon hazard index

th th th REL h i T bl 5rather than the RELs shown in Table 5.

But wait, the units have changed again! Now they are in ppmv. Same numerical value, but different units from both Table 1 & Table 5.

22

In the text above, the consultant lists the BTEX concentrations as ppmv. The values of 0.3 ppmv exceeds CA standards by approximately 250 times.

23

But in Table 9, the units are listed as ppbv. If correct, the measured values are BELOW allowed values and the site passes the VI assessment!

24

Although it is acknowledged that ambient levels are the likely cause of these elevated levels; another round of indoor sampling is recommended per the guidance document.

All of this work being done & more proposed, even though the initial screening for this site documented that there would not be a risk to indoor air from the closest soil gas samples.

25

26