MAKING SENSE OF 8260 AND TO-15 RESULTS FROM SOIL GAS SAMPLING EVENTS: WHY THE DISCREPENCY?
This is a good question and we will attempt to explain some of the problems with differing results that come from the same sampling location.
First, one of the largest sources of error is Lab-to-Lab variations. EPA, in order to determine quality control limits for each method, sends out samples to at least 20 different labs. Each lab is sent the same samples and asked to analyze these samples as a routine analysis. When the results come back, the EPA does a statistical analysis of the data and reports expected quality control limits. Most of this work is performed through the Contract Laboratory Program (CLP) as part of the Superfund regulatory mandate. EPA usually uses the 2 standard deviation rule when setting up quality control ranges. This includes, by definition, 95% of the values obtained. Listed below are examples of some of the acceptable ranges for particular analytes:
| Analyte | Acceptance Range | Reference |
| PCB-1260 | 8-127% | 40CFR136 (Method 608) |
| Benzene | 37-151% | 40CFR136 (Method 624) |
| Napthalene | 21-133% | 40CFR136 (Method 625) |
| Toluene | 76-125% | CLP (Volatiles) |
| Phenol | 26-90% | CLP (Semi-Volatiles) |
As evident, the ranges can be quite broad. It should be noted that CLP is notoriously stricter on labs than the general regulating community.
Another area worth highlighting involves proficiency testing and all certified laboratories in California must participate in this process. In order to become certified for a given method/analyte the laboratory must analyze proficiency samples (PE samples). These PE samples are sent out to hundreds of environmental testing labs and then processed to determine acceptable ranges for each analyte. The table below gives a sampling of the ranges obtained in the most recent PE test results.
| Analyte | Acceptance Range | Method |
| Gasoline | 156-1620mg/Kg | LUFT/8260/8015 |
| Diesel | 363-1560mg/Kg | EPA 8015 |
| Benzene | 49.5-115mg/Kg | EPA 8260 |
| 1,4-Dichlorobenzene | 37.9-156mg/Kg | EPA 8260 |
| Tetrachloroethylene | 16.1-53.4mg/Kg | EPA 8260 |
The ranges listed above are by no means the worst with many variations in the range encompassing more than an order of magnitude. As an example of how bad reproducibility is Lab-to-Lab, let’s take gasoline as an example. If lab A got a number on the same sample for gasoline of 156 mg/Kg and lab B got 1620 mg/Kg, then the results from these two labs would be acceptable. Consequently, this is why it is highly recommended not to change labs when comparing trends in ground water data.
As a rule-of-thumb, lab-to-lab variations range from 60-140%.
The second source of analytical error/differences occurs with the methods. EPA 8260 uses purge and trap extraction/concentration of sample analytes where water is always added to the purge vial prior to extraction. Standards for this method are liquid. EPA TO-15, on the other hand, uses gas standards and traps the gases at the injection port using cryo-focusing with liquid nitrogen as a cryogen. For example, the TO-15 procedure is more adept at analyzing compounds that have high water solubilities since they are difficult to purge out of water using the EPA 8260 procedure. There is no question that there are significant differences between the two methods especially for those compounds with high vapor pressures and low water solubilities. How much of a difference between the two methods depends on the laboratory doing each of the methods. If the same laboratory is performing the two tests, you can expect numbers that are closer since you will eliminate lab-to-lab variations. These same labs typically perform sample collection and handling similarly.
Many examples exist of values differing between methods. Examples where numbers differ by more than an order of magnitude between the two methods are not uncommon. Typically, method-to-method variations will range between 20-100% but can differ over at least an order of magnitude.
Third, the new DTSC/RWQCB guidelines provide guidance on how to take soil gas samples. Even with the high degree of detail provided in the Soil Gas Advisory, sample collection in the field will vary considerably from person-to-person and lab-to-lab. Soil gas samples should never be collected by the driller!! Consistency in sampling by the individual are important especially when collecting duplicates and if one intends to collect samples at different times during the day. Barometric pressure, humidity, purge volumes, purging device, sample containers, and many other parameters affect results when sampling is considered.
Fourth, container differences pose considerable potential differences since the containers used to collect soil gas samples differ greatly. Glass gas-tight syringes are the most popular for EPA 8260 analyses where the syringe or a sampling pump is used to purge the probe before collection. Glass bulbs can also be used to collect soil gas samples. Syringes provide the best collection device since the collector can accurately determine the volume removed from the probe and its adaptability to in-line connection to a Magnehelic vacuum gauge. In-line connection of a Magnehelic vacuum gauge to the sampling train provides immediate feedback of the vacuum during purging of the tubing apparatus and during the sample collection. Accordingly, vacuum monitoring via the Magnehelic ensures consistency and adherence to sampling protocols during the sampling event. What's more, the use of the Magnehelic allows for differentiation between “no-flow” and “low-flow” sampling events. This is a crucial point that cannot be emphasized enough. “No-flow” equates no sample collected and sample probe is either abandoned or re-set (either case is not desirable). On the other hand, “low-flow” can be accurately assessed by the vacuum reading on the Magnehelic and given vacuum does not exceed 100 inches of water a sample can still be collected (typically “low-flow” situations require a low flow rate, e.g., 25-50cc/min).
Soil gas investigations with Summa canister collection followed by stationary laboratory analysis does not routinely have the advantage of providing a purge test. Summa canisters must be collected completely; that is, if the Summa can is a 6-liter can then 6 liters must be collected. Although it is possible to collect different volumes in different sized Summa canisters to determine the optimum purge volume, one has to wait for the Summas to be analyzed before proceeding with the rest of the soil gas surveys. This will take days or weeks. One consequence of Summa canister use is that the Summa may be collecting too much sample. For instance, if the lab has determined that the optimum purge volume is 3 purges, a 5 foot probe will require 400-500cc of purging. If the stationary lab sends a 6-liter Summa (or even a 1-liter can), you will be collecting 12 to 15 times (or 2 to 2.5 times) the optimum volume required for that site depth. The likelihood that the results from a 6-liter (or l-liter) Summa sample will match a 400-500cc EPA 8260 sample is not very good. Another problem associated with Summa canister usage is attaching the Summa canister to the sampling probe. Improper attachment to the probe will cause leaks and/or plugs when collecting a Summa canister sample. Also, if the sampler does not purge the probe the same way as the EPA 8260 sample before taking a Summa canister, they will introduce ambient air into the Summa can and will not collect the optimum purge volume.
Differences in sampling containers can vary by as much as 2 orders of magnitude especially when comparing Summa cans with glass syringes or bulbs. Typically, different collection containers will give differences between 25-100%. Additional potential differences between the two methods are within lab variations (typically + 15%) and weather conditions (maximum + 10%).
As you can see, using two different methods by two different laboratories will introduce large potential differences in laboratory results. It’s a wonder that most of the results are as close as they are.
The use of an on-site mobile laboratory will tend to minimize most of the differences described above. If problems such as leaks occur, the mobile laboratory allows the consultant to immediately determine the problem and re-run the sample the same day. In addition, optimum purge volumes can be obtained as well as step outs for additional data/probes. Coupled with the fact that drillers can re-drill plugged probes, the consultant has much more flexibility in their sampling plan and can provide a more cost effective report to their client. High quality mobile laboratories provide many advantages over sending soil gas samples to stationary labs.
