(621a) The Impact of Trace Ammonia Impurities on the Utilization of Renewable Natural Gas

Renewable natural gas (RNG), produced from biogas upgrading, is an important alternative to fossil fuels for power generation. Unfortunately, RNG contains several trace contaminants. One such contaminant is NH₃, of particular concern for RNG produced from farming operations. The presence of NH₃ in RNG, particularly if it is injected into the natural gas (NG) pipeline network, can have serious consequences. These can range from damage to components of the NG infrastructure, to corrosion of analytical and end-user equipment and added pollutant formation during combustion. The research in this paper addresses all such impacts.

In the area of pollutant emissions, three common end-user equipment were tested: An internal combustion engine, a household tankless water heater and a wall furnace. They were all operated with NG injected with NH₃ concentrations ranging from 0 to 500 ppm_v. Emissions studied included unburned hydrocarbons (UHC), CO, and nitrogen oxides (NO_X), the latter being of particular concern. The presence of NH₃ resulted in increased NO_X emissions for all equipment. For some of the equipment, the relationship between the amount of NO_X formed and the NH₃ concentration in the RNG was direct and quantitative, meaning that for every molecule of NH₃ fed to the equipment, one molecule of NO_X was produced. For others, a lower quantity of NO_X was formed to that corresponding to complete conversion, underscoring the impact that the prevailing conditions in the combustion chamber has on the pollutant formation processes.

We also investigated NH₃-induced corrosion in materials relevant to the NG infrastructure with the goal to identify various (bulk and surface) species potentially forming, their rates of formation, and their dependence on the operating conditions. The tested materials included components made of steel and brass, and a variety of polymer gaskets. They were exposed to both dry and humidified (RH>90%) NH₃/ NG mixtures with 0-500 ppm_v of NH₃ for periods, under each concentration, ranging from 400-1200 hours. NH₃-induced changes in the materials were identified via direct weight change measurements and optical observations, SEM-EDS, ICP-OES and microhardness testing. For the metal components, corrosion phenomena were observed, particularly for the higher NH₃ concentrations under humidified conditions.

We also studied the impacts of NH₃ in the NG may have on Gas Chromatographs (GC) used to analyze the gas for customer billing. Two issues of concern were: (i) interference in the measurements due to the presence of NH₃; (ii) corrosion of GC internals due to its long-term exposure to NH₃. Two different experiments were conducted to evaluate each of these concerns. In both experiments, the GC was gradually exposed to NG containing various NH₃ concentrations ranging from 30 ppm_v to 200 ppm_v. For experiments probing interference, the duration of experiments was rather short, typically, consisting of 50-70 measurements (each measurement lasting 5 min., 14 sec) for each NH₃ concentration studied. For the experiments to study potential instrument corrosion, each cycle of experiments for a given NH₃ concentration lasted about a week. Two different experimental protocols were utilized: During the early phase of the study, for the first group of experiments, the instrument was calibrated at the beginning of the experimental run using a standard calibration gas, but for the whole series of experiments it was not calibrated again. The reason for doing that was to observe any cumulative effects that continuous exposure to NH₃ may have on the instrument. In the second group of experiments, we followed the normal operation protocol for the instrument when used in the field by re-calibrating the instrument daily. For both the interference and the corrosion experiments, the presence of NH₃ has a detrimental impact, more so, as expected, for the long-term (corrosion) experiments rather than the short-term (interference) experiments. The impact was significantly more pronounced for the experiments without re-calibration, during which at some point, and before the planned completion of experiments, the GC stopped functioning all together. The daily re-calibration using standard gas had a good benefit, with the NH₃ impact on the instrument still being present but not to the extent encountered in the runs without re-calibration.

The results of the study are concerning, indicating the need for thorough purification of the RNG prior to its potential injection into the NG network.