Triangle NoTES

 

Mid-June 2018

 

There have been some ongoing discussions concerning our long established practice of correcting for air intrusion in the Method 25-C samples using whichever of the oxygen or nitrogen fractions gave the minimal correction based on the ratio in ambient air. This was based on guidance received when I brought the problem to the attention of the EMC in 2002. The more recent change to the method only allows for the correction of a valid sample using oxygen concentrations if the nitrogen concentration is above 20%, which retains the problem for all of the samples for which the nitrogen is less than 20%.

 

I have calculated the retained bias by taking a random landfill sample's field data to calculate sample volume, etc. and arbitrarily set the NMOC concentration at the analyzer to 100 ppmC. I then calculated the results based on 100 ppm of oxygen and a nitrogen concentration 1 ppm over the 20% level. This would allow the oxygen to be used in calculating the correction factor for air, which for this example gives an NMOC concentration of 754 ppmC. I then recalculated the results using a nitrogen concentration of 1 ppm under the 20% level, which would allow only the nitrogen concentration to be used to correct for the air intrusion. In this example the corrected result for NMOC was now 1011 ppmC. This is a 35% positive bias from one correction to the next, which results from a difference of literally only 2 ppm nitrogen in the sample.

 

I have a spreadsheet which can be emailed to anyone interested in looking at the way the correction affects the results in various different situations.

 

This is why we use the oxygen concentration in the majority of the sample calculations. The oxygen to nitrogen ratio generally shows an excess of nitrogen as compared to ambient air, which indicates the source for the nitrogen is not ambient air intrusion during sampling, which is the basis for the acceptance criteria of 5% oxygen or 20% nitrogen for a valid sample. This also means the correction should be for ambient air intrusion and not for just the nitrogen concentration and a proportional concentration of oxygen for that level of nitrogen.

 

I believe the cutoff at 20% for nitrogen is probably related to the worst case samples I presented to the EMC in 2002, because in those cases the nitrogen was in the 40% to 50% ranges and there was less than 5% oxygen. I had assumed that any change would indicate the ratio of ambient air, as we had been doing, for all samples not just those with concentrations of nitrogen above 20%. I knew there was still an error when I saw the change, but I assumed the prior guidance would suffice so I did not push the issue. I have recently been told that any change such as this negates the prior guidance, which does make it a problem now.

 

This issue goes back to the point where I first read the proposed Method 25-C and noted the nitrogen correction in the formula was written as if there was no oxygen in the air. I think the initial concept may have been to combine the two fractions, which would have added another layer of problems due to the disproportionate concentration levels in the samples compared to ambient air, but would have explained the initial formula.

 

 

One of the initial concerns was the potential dilution of the landfill gas sample with ambient air during sampling, which would have created a negative bias in the reporting of the NMOC concentrations. When the Method 3-C/25-C combination was first presented, the allowable nitrogen concentration was, I believe, 1% to be valid and there was a correction factor calculation to apply to the reported NMOC concentration. Once the formula was corrected to properly account for the nitrogen fraction, this was to be used for all such projects. The 1% level was not readily attainable in field testing and over the years was incrementally increased to the current <20% nitrogen or <5% oxygen criteria.

 

This is important in that the initial EPA position acknowledged that a 1% air intrusion, as evidenced by the nitrogen concentration, created a bias significant enough to warrant such a correction. This assumption of nitrogen only coming from ambient air intrusion proved to be incorrect once the field data started showing otherwise. After a particularly disproportionate set of samples, I presented my concerns to the EMC concerning the accuracy of using only nitrogen to determine whether a sample was valid. This was based on samples in which the nitrogen content measured as high as ~40% but with minimal oxygen levels, which clearly indicated no ambient air infiltration had occurred during sampling. This later led to the current criteria of <20% nitrogen or <5% oxygen for a sample to be considered valid.

 

This acknowledgment that landfill samples do not always contain proportionate levels of oxygen and nitrogen led to questions from sources concerning the nitrogen correction on valid samples where there clearly was not ambient air for that level of correction. This created a significant positive bias against the sources which encountered this situation. Since there had been acknowledgment of the disproportionate levels of oxygen and nitrogen in the validity criteria and acknowledgment that a 1% nitrogen level created a negative bias significant enough to warrant a correction, there was a strong opposition to allowing a 1% or greater positive bias considering this could trigger action levels which could be expensive for those sources.

 

I again was in contact with the EPA to illustrate the problems with performing just the nitrogen concentration correction and the industry concerns relating to it. I did receive an email in return indicating that my proposal for an either/or correction was a logical and acceptable solution. After about a decade, the current adjustment was included, but it was worded as such to only deal with valid samples with nitrogen concentrations above 20% and to only then adjust for oxygen. This ignored the majority of the samples which have a nitrogen concentration less than 20% but still generally higher in proportion to the oxygen concentration. When these samples are adjusted to account for the ambient air intrusion using the nitrogen concentration, the NMOC results are artificially inflated creating a positive bias. Few of the Tier II samples have a nitrogen to oxygen level lower than ambient air. I have reviewed several Tier II projects and with the new criteria, 56% of the samples had some level of positive bias when a crude 4:1 ratio for ambient air was applied, 30% would have had no bias regardless, and 14% were no longer biased due to the recent change in the application of the oxygen correction.

 

The existing either/or adjustment only mitigates the most egregious levels of bias, not the more usual, but still significant levels of bias. The correction biases for 100 ppm oxygen with 19.99% nitrogen compared to the correction for 4.99% with 40% nitrogen are both significant, but only one is covered under the change to the method and the other is currently accepted. What is often missed is that for every 4% nitrogen for which the NMOC concentration is corrected there is also a 1% correction for oxygen even if there is less oxygen in the sample. This disproportionate ratio of the two major components of ambient air means that only one will correct to an air free basis and the other will create a comparative bias. This is why a pure either/or allowance for the correction of ambient air based on whichever corrects only for ambient air should be the standard for the method.

 

The method indicates that the intent is to correct for ambient air intrusion, but the stated correction criteria creates a relative positive bias if the nitrogen level is below 20% and not above that concentration. This is contrary to the initial acknowledgment that a 1% nitrogen level created a significant bias. The only difference between the two positions seems to be the direction of that bias. It appears that currently any bias in favor of the source is a problem but bias against the source is acceptable if it is only significant and not egregiously significant.

 

My concern is that the results we report are the most accurate and applicable as possible and in that regard I will continue to try to ensure there are no issues which are possible to mitigate.

 

Currently, the EMC has suggested that I submit a request for an alternative method. This concerns me in that I, like most people, would consider an alternative method as one which uses some type of different process to perform the sampling or analysis but which gives an equivalent result to the original method. In this case the alternative method will never give the same results as the original method because the applied correction will be different for most of the samples to which it would be applied. In effect the original method would not be an appropriate substitution for the alternative method, but the alternative method would always be an appropriate substitution for the original method. This seems very counterintuitive for a standard reference method compared with an alternative method to me. I know of other instances where alternative methods were appropriate while others were not appropriate options for this very reason. This type of situation raises the real question concerning other alternative methods and whether they should actually be the reference method and how would anyone know which would be the appropriate choice in compliance situations. Also, in the case of requiring accreditation for compliance purposes, the accreditation would not be for the original method, but the more correct alternative method. If in a case such as this, the results are significantly different depending upon which method was used, it would be very confusing to determine which met the desired compliance requirements. In a case where the results for the standard reference method were less correct than the alternative method, would the permits need to specify the alternative rather than the standard method? I am not really expecting an alternative method solution to work just for these reasons.

 

Following are the specific sections of Method 25-C which are applicable to the correction for ambient air intrusion.

 

8.4.2 Use Method 3C to determine the percent N2 in each cylinder. The presence of N2 indicates either infiltration of ambient air into the landfill gas sample or an inappropriate testing site has been chosen where anaerobic decomposition has not begun. The landfill gas sample is acceptable if the concentration of N2 is less than 20 percent. Alternatively, Method 3C may be used to determine the oxygen content of each cylinder as an air infiltration test. With this option, the oxygen content of each cylinder must be less than 5 percent.


 

    1. Miscellaneous Quality Control Measures.

Quality control measure:

8.4.2 Verify that landfill gas sample contains less than 20 percent N2 or 5 percent O2

Effect:

Ensures that ambient air was not drawn into the landfill gas sample and gas was sampled from an appropriate location.

 

 

12.5 You must correct the NMOC Concentration for the concentration of nitrogen or oxygen based on which gas or gases passes the requirements in section 9.1.


 

      1. NMOC Concentration with nitrogen correction. Use Equation 25C-4 to calculate the concentration of NMOC for each sample tank when the nitrogen concentration is less than 20 percent. New :(If COX/CN2 > 21/78 then use the nitrogen correction formula)

 

      1. NMOC Concentration with oxygen correction. Use Equation 25C-5 to calculate the concentration of NMOC for each sample tank if the landfill gas oxygen is less than 5 percent and the landfill gas nitrogen concentration is greater than 20 percent.

New : (If COX/CN2 < 21/78 then use the oxygen correction formula)

 

 

 

 

 

 

 

Wayne Stollings

Triangle Environmental Services, Inc.

Wstollings@aol.com

 

 

(919) 361-2890 (800) 367-4862 FAX: (919) 361-3474