I thought it might be time to revisit some of the history, as I understand it, of Method 25 and what is actually viable using the method, since those FAQs seem to have disappeared from the EMC website.
As some of you probably already know, there was a connection between Method 5 and Method 25. This created the theoretical division at the time between particulates and VOCs. Method 25 lists a potential interference of particulate matter, which is the reason for the heated filter. The initial thought was the use of the same filter as Method 5 at the same temperature of ~250 F. This filter requirement was never incorporated into the method, so for many years filters were used which were not compliant with the requirements of Method 5. This concept created the potential for a negative bias when Method 25 was used in inlet monitoring, which was discovered by accident in a few test programs when the filters were changed to the Method 5 compliant filters. Since very small particulates and aerosols are being introduced into the control device, their exclusion lowers the measured loading so it is not considered when the outlet concentration is measured. If the particulates/aerosols are unaffected or are all completely destroyed there is no problem, but if there is any reduction in size that would allow them to be included in the outlet concentration measurement there can be a significant bias created. It is clearly appropriate to exclude particulates in the measurement of outlet concentrations, but the aerosols may be an issue in some cases. The inlet, however, should be given a much wider range of use where a destruction efficiency must be calculated.
The other interference listed for Method 25 is CO2 and H2O. The method specifies that the bias created is insignificant when the product of the concentration of both compounds is 100 or less. If the product is above 100 the bias may be significant. The assumption was that a matrix of ice could trap sample gas containing CO2, which would not be completely flushed from the trap during the first stage of the recovery. The entrapped CO2 would be counted as NMOC, thus creating that feared positive bias. In my discussions with some of the people involved with this determination, the actual measured bias percentage in the research was really not that significant. This was somewhat confirmed when I did some rough calculations on how much the CO2 laden sample gas would affect the sample at various volumes being trapped. The amount of sample gas required to have any significant impact was a large percentage of the volume of the entire trap. The other aspect of the moisture concentration is the effect on sampling. The smaller tubing going into the larger tubing at the end of the trap can cause problems with ice plugs stopping the sample flow. Once the flow stops the trap must be warmed enough at that spot to allow the ice to melt and flow to resume, or a second, or third or fourth trap used to replace the plugged trap. There is much more of an issue with the multiple traps in sequence due to the problems with calculating an accurate concentration for each trap. This would be accomplished by using only the volume sampled for that specific trap and combining them all using a relative average, so the field expedient solution is to warm the trap. This solution works, but there is also a potential for warming the trap too much thereby allowing some of the organics which should remain in the trap to migrate into the tank. There is always a slight amount of NMOC from the condensible fraction making it through to the tank due to the vapor pressure, but warming the trap can greatly increase this amount. The impact of this migration is unknown because there are many variables for each project, but the potential for a negative bias seems greater over all.
The potential positive bias played a part in the U.S. EPA Guideline Document dated April 4, 1995, which essentially set the division for use between Method 25 and 25A at 50 ppmC as far as most people were concerned. It required Method 25 for DE afterburner compliance and allowed for 25A to be used when the outlet concentrations were at or below 50 ppmC. If the concentration was above 50 ppmC using Method 25A, it was not the correct method for the application. Costs were a consideration, but the primary complaint seemed to revolve around the positive bias at lower levels being more significant. Of course any bias is going to be more significant at lower levels, but the question is how significant that bias is. There were issues in the method prior to the February 1988 changes, but even after that time the new method was tied more to a 3.6 L sample volume than not. The minimum sample rate was 60 cc/min for an hour and with the sample tank sizes of 4 L and 4.5L being most common for much of this time that was the basis for the volume of sample taken. There was some movement toward a 6L sample tank prior to our use of the 8.3 L sample tank. Since the sample trap operates similarly to a TO-14 sample concentrator in a field application, the larger the volume sampled the less the impact of any bias. In the decades since this Guideline document was written, we have reported almost identical sample concentrations in sets of triplicate samples below 20 ppmC and reported samples in the range of 10 ppmC. We have also passed audits provided to field samplers under the U.S. EPA audit program at 50 ppmC. This was accomplished by reporting the correct concentration within the +20% allowable range. This means there was less than a 10 ppmC deviation from the sample being collected in the field, through the recovery process, and ending with the analysis and final calculations. We cannot state the reported concentrations are accurate at those levels due to the potential biases, but we can be sure that the concentrations reported for the sample are not greater because of those same potential biases. Thus, if the reported concentration is 17 ppmC, we are confident that it is not 18 ppmC or 19 ppmC, but it may actually be 16 ppmC or 15 ppmC due to the bias potential.
The potential for bias is not limited to the moisture and CO2 content, however. This is a carbon counting method which recovers the trap fraction at a temperature of 200 C. Thus, any carbon which makes it into the sample stream and releases any carbon at that temperature can result in a bias. For example pump oil, which is a heavy molecular weight hydrocarbon, does not impact the tank fraction due to the low vapor pressure. It can, however, be an issue with the trap fraction since the 200 C temperature will cause some volatilization, which will in turn be oxidized and counted as part of the sample NMOC. Over the years we have seen traps which have had what appeared to be pocket lint and Teflon tape in the cap and the end of the tubing at the fitting. Both of these can add carbon to the trap sample just like any amount of pump or skin oil can. This is sometimes why it is so difficult to use the method in areas where there is a higher background of contaminants which cannot all be removed. One memorable set of samples came from a paper facility making toilet paper products. There were small bits of toilet paper waste over everything including the sample tanks, the sample traps, and the sample consoles. Any amount of something like that being introduced into the trap will create a huge bias. We have seen directly and have heard of other situations where customer owned sample consoles were used for testing after being cleaned and still contaminated the samples. We do not often see a console which does not pass the QA/QC checks after being cleaned now, but it has certainly been seen several times over the years. It does not matter if it is only a one in a million chance of contamination if your project is that one in a million. The potential for contamination during the field sampling is understood by many of the regulatory bodies and also supported by an impressive data set provided to the Emission Measurement Center by NJDEP years ago.
There is a lot which can be done with the method, but it must be done correctly from start to finish.
Triangle Environmental Services, Inc.
P.O. Box 13294 122 US Hwy 70 E
Research Triangle Park, NC 27709 Hillsborough, NC 27278
(919) 361-2890 (800) 367-4862 Fax: (919) 361-3474