Recently I was approached with some questions concerning landfill gas (LFG) analysis, which were based on what appeared to be logical positions, but which were less so once the larger picture was viewed.
The first question concerned the use of just the methane analysis from Method 3-C for the calculation of the BTU heating values for the gas. It appeared that a more comprehensive approach might give a more accurate result. This was the subject of discussion very early in the evolution of the Method 3-C/25-C program for LFG. As there is a widely varied potential composition of combustible gases in each LFG sample, the practical solution was initially to use the carbon from the Method 25-C analysis converted to hexane rather than to analyze for the ever increasing number of trace compounds to try to add each to the calculation of heating value. However once the data was reviewed, the impact of the trace gases on the reported BTU calculations was found to be negligible. The normal variation of the methane in the LFG was as much or more than the added hydrocarbons. Thus, for simplicity the methane concentration from the Method 3-C became the most effective method for this calculation.
To better illustrate this situation, I pulled several old reports at random to compare the reported BTU heating values with the potential change from adding the NMOC as hexane to the total. I would also rather use some actual examples than try to make generalizations from memory, thus these are from actual reports but with insufficient information to identify any particular report.
I picked 5 reports with a Gross BTU value per cubic foot of 493.2 to 529.4. The Gross BTU value calculated for the NMOC concentration corrected to an air free basis and converted to hexane gave a range of 0.5 to 2 for these samples. Since the NMOC and methane concentrations are not linked, the upper and lower ranges are not necessarily the same samples. Even so the use of the lowest methane BTU value and the highest NMOC BTU value do not indicate a very significant variation when all factors are considered. The reported concentration of methane is calculated from the triplicate analysis of the sample using the average concentration as the most accurate. The methane concentration error potential for these samples ranged from 111 ppm to 302 ppm, which would equate to a BTU value range of 0.1 to 0.3 with all other variables remaining the same. I say this because another significant variable is moisture content, which is assumed to be saturation at the ambient pressure and temperature of the final reading. This is based on the assumption the LFG is hotter and therefore contains more moisture than the ambient conditions would allow. The fact we use dry helium to pre-charge the canister to ~40% of the total volume should prevent any actual condensation of water, but since there is no good method to measure the moisture in the LFG it is the best assumption to date. These five samples had the oxygen, nitrogen, carbon dioxide, and methane totaled after the adjustment to the dry basis. The range of the totals were 96.0% to 100.2%, which is a fairly common range for such samples.
The errors in measurement for the pressures, temperatures, volumes, raw analytical data, the corrections for air infiltration and correction for water content can combine to easily cause the calculated BTU value to exceed the additional NMOC value for most samples. This means the added cost of more analyses really would not actually improve the results more than just appear to improve them. This does not consider the variation of the methane concentration during the same day of the sampling or between different days of sampling, which we see from sources which are measured on a regular monthly or quarterly basis.
The other question involved a carbon balance across a device and what compounds would be likely to be included. The assumption was that the NMOC and carbon monoxide would be necessary for an accurate calculation. This also seems logical until the measurement errors are examined.
I do not remember actually seeing carbon monoxide in any LFG samples but it is possible I am not remembering correctly. The reports I pulled looking for reported concentrations all listed carbon dioxide as being below our reporting limits. A few years ago a state agency was trying to set a standard to use for determining whether there was the potential for a fire in a landfill based on the carbon monoxide concentration in the LFG sample. The lower limit they were discussing was 100 ppm, which is much higher than our report limit, so I will use that as my base. The NMOC concentration is also highly variable between samples. For example, the range of NMOC for the sets I pulled ranged from 610 ppmC to 2835 ppmC after the correction to an air free basis. The uncorrected concentrations would be more accurate for a carbon balance, but that is not as common.
The concentration error for the five samples for carbon dioxide ranged from 79 ppm to 222 ppm again based on the adjustment to a dry basis per the method calculations. This means the error for the average reported concentration of carbon for just the carbon dioxide and methane ( 111 ppm to 302 ppm) alone is almost as much as the lowest range of the NMOC, which is itself adjusted up for the subtraction of air from the sample. If it was taken on the as analyzed basis the measurement error would exceed the lowest range of NMOC and the highest range of carbon monoxide combined. The NMOC reported uncorrected for air would be an expected addition to the carbon balance given the cost to value comparison, but the additional analysis of carbon monoxide would seem to be in the diminishing returns for investment category for most situations.
I have always tried to consider whether there was a value to a client in the expense of an analysis. I have occasionally talked myself out of work, but I have always viewed that as an investment in the partnership between all of the parties involved, The ultimate goal is to provide the best and most reasonable data for the most reasonable costs. The return for investment determination is where I see the regulated community and the regulatory community working together to have that best data for the most reasonable costs possible. It also requires an occasional view of what the larger picture is for the data and whether it is being improved upon or just changed by whatever is being considered. This is why we no longer provide the carbon dioxide, methane, and the optional carbon moxoxide concentrations with our Method 25-C report. The addition of those compounds caused more potential harm to the data than improvement.
To condense this into a couple of easy positions. The use of Method 3-C to determine the heating value for LFG should give the better data for the better cost in most situations. The use of both the Method 3-C for the fixed gases and the Method 25-C for NMOC in a carbon balance calculation should give the better data for the better cost in most situations as long as no correction for air infiltration is performed.
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