** Triangle No****TES**

February 2019

To continue with the discussion on the 'Ultimate Analysis' of landfill gas with some of the feedback and what we have determined in the interim.

First, it is understood that some of the elements would not be included in the calculations. The trace sulfur and chlorine, for example. There would also not be an inclusion for the hydrogen and oxygen from the organic compounds reported in the Method 25-C analysis. While there can be even more analyses to determine these components, to a greater degree it seems to be a strong potential for the diminishing return on expenditures.

To better determine what the results would resemble for a landfill gas report, I did a couple of groups of calculations based on a set of three samples which showed consistent results over all of the analysis.

On the wet basis, the mid-range sample had 35.15% CO2, 47.36% CH4 , 12.31% N2, 2.14% O2,, and 0.07% NMOC which left a calculated 2.98% H2O to bring everything to 100%. This translated to a weight basis of 55.76% CO2, 27.38% CH4 , 12.43% N2, 2.14% O2,, 1.93% H2O and 0.03% NMOC using the molecular weight of the components.

Converting this to the elemental components and using the atomic weight of the individual atoms for a weight calculation created a difference just from rounding error. The MW of CH4 is different from the accepted weight of four hydrogen atoms and one carbon atom due solely to rounding. The difference is 0.01 lower for the MW of CH4. With the amount of CH4 in the sample, the difference total is 0.02% higher using the atomic weight of the atoms.

Converting the results to a elemental level and taking the analytical variation between the three analyses of each sample the results are 35.74 +0.01% C, 44.72 +0.02% O, 12.43 +0.01% N, and a constant 7.11% H, which does include water vapor. Some of the references I have found for the solid fuel reports include the water content in the elemental percentage, while others list the water separately from the elemental percentage. The difference seemed to be related to the determination of BTU value based on the elemental percentages. Since the BTU value can more easily be determined by the molecular components in the gas samples and the elemental approach would be mainly for gas density, the water being included in the elemental percentage seem more reasonable to me.

If we do not include the Method 25-C results and use only the Method 3-C as seems to have been indicated in the Subpart D reference to fixed gas analysis, the results become 35.71% C, 44.75% O, 12.43% N, and 7.12% H. This loss of NMOC as carbon does not seem to be very significant when compared to the rounding and analytical uncertainty variations and especially when the costs are included. It does, however, help to offset the other elements not included due to even higher analytical costs.

The percentage breakdown of the molecular components by volume and by weight both including and excluding the NMOC carbon portion:

CO2

CH4

N2

O2

NMOC

H2O - by

subtraction

35.15%

47.36%

12.31%

2.14%

0.07%

2.98%

By vol

55.76%

27.38%

12.43%

2.47%

0.03%

1.93%

By wt

35.15%

47.36%

12.31%

2.14%

NA

3.04%

By vol -NMOC

55.75%

27.38%

12.43%

2.47%

NA

1.98%

By wt -NMOC

The percentage breakdown of the atomic components:

C

O

N

H

35.74%

44.72%

12.43%

7.11%

Incl NMOC carbon

+0.01%

+0.02%

+0.01%

+0.00%

Analytical variation

35.71%

44.75%

12.43%

7.11%

-NMOC carbon

Note: The MW of CH4 is 16.04 as opposed to 16.05 when the rounded atomic weights are used. This makes a 0.02% increase in the atomic versus molecular weight comparison.

Now the information I received in feedback on projects which had used Method 3-C for a mass or density calculation can be discussed. There were two projects discussed and one was more of a biogas than landfill gas location.

The first project was one in which the correction for moisture resulted in a total in excess of 100% for the three samples involved. The solution was to normalize the reported concentrations to 100%, which seemed like a good choice at the time. In our recent discussion I pointed out this meant the measured concentrations were adjusted by an assumed moisture content and then normalized to 100% thereby leaving the moisture at the assumed level and adjusting the measured components accordingly. Fortunately, leaving the measured components at the wet basis and assuming the adjustment for moisture would take that to 100% made an insignificant change to the calculations which had been previously used.

The second project was one in which the correction for moisture was fairly accurate given the totals were near 100% with a close over-and-under range. For this project the components were averaged over the three runs and those averages were used in the final calculations. This also included the average of the three analyses of NMOC as carbon in the totals. The average total did not exceed 100% but was very close so normalization was not considered. This is almost the same process which would be used in totaling the components as wet basis and then considering the moisture as being the remaining percentage. The averaging of the three runs would be a very good addition to the calculations as well.

From what has been used in the field already and what I have seen in these calculations of the single test run, there is little impact from the uncertainty of the analysis or the trace compounds which are not included. The conversion to the elemental level adds a slight positive bias due to the rounding issues, which could be considered as an offset to the lost elements. Whether that offset is sufficient or too much will be something a regulatory entity will have to determine if this process is to be used to determine the fuel density. The addition of the NMOC from the Method 25-C does not seem to be critical nor significant, but it is an easy addition which arguably makes the total calculations more accurate. Given the alternative here would be for the unreported percentage carbon from NMOC to be replaced by moisture, the density of water is greater than that of carbon if that were used. If even an ordinary hydrocarbon, such as hexane, were chosen to be used in place of the NMOC only the dry basis density would change as the moisture density would be greater than the hydrocarbons.

How does this all shake out? Just from the cost to benefit comparison it would be like this:

Method 3-C measured concentrations wet basis with only moisture assumed to make the difference to 100% total weight. The lowest cost approach and seems to give ~99+% of the density of the gas .

Methods 3-C and 25-C measured concentrations wet basis with only moisture assumed to make the difference to 100% total weight and converting the NMOC to at least hexane for a slightly better assumption. Higher cost and maybe only the perception of greater accuracy. This approach still seems to give ~99+% of the density of the gas if only just slightly higher.

Methods 3-C, 25-C, and 16 for Total Reduced Sulfurs wet basis with only moisture assumed to make the difference to 100%. Much higher cost and maybe only the perception of greater accuracy. This approach still seems to give ~99+% of the density of the gas.

Any other analytical methods will increase the costs, but really have little impact on the gas density given the insignificant percentages involved. Even the cumulative effect does not seem to be noticeable given the calculation for the NMOC compared to the water variation.

The change to density for the dry basis will be the largest changes, as the moisture density would offset much of these minor additions for gas density on the wet basis.

Wayne Stollings

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