Interferences caused by the biogeochemical methane cycle in peats during the assessment of abandoned oil wells

Abstract. In the global effort to reduce anthropogenic methane emissions, millions of abandoned oil and gas wells are suspected to be prominent, although thus far often overlooked, methane sources. Recent studies have highlighted the hundreds of thousands of undocumented abandoned wells in North America as major methane sources, sometimes emitting up to several tons of methane per year. In Germany, approximately 25 000 abandoned wells have been described; these wells have been well documented, and the data are publicly available. Here, we present a methodological approach to assess emissions, particularly methane, from cut and buried abandoned wells, which are typical of wells in Germany. We sampled eight oil wells in a peat-rich environment, with four wells in a forest (referred to as Forest), three wells at an active peat extraction site (referred to as Peat), and one well in a meadow (referred to as Meadow). All three areas are underlain by peat. At each site, we sampled a 30 m × 30 m grid and a corresponding 20 m × 20 m reference grid. Three of the eight wells and reference sites exhibited net methane emissions. In each case, the reference sites emitted more methane than the respective well site, with the highest net emission (∼ 110 nmol CH4 m−2 s−1) observed at one of these reference sites. All methane-emitting sites were located within the active peat extraction area. Detailed soil gas characterization revealed no methane-to-ethane or methane-to-propane ratios typical of reservoir gas; instead, it showed a typical biogenic composition and isotopic signature (mean δ13C-CH4 of −63 ‰). Thus, the escaping methane did not originate from the abandoned wells nor the associated oil reservoir. Furthermore, isotopic signatures of methane and carbon dioxide suggest that the methane from the peat extraction site was produced by acetoclastic methanogens, whereas the methane at the Meadow site was produced by hydrogenotrophic methanogens. However, our genetic analysis showed that both types of methanogens were present at both sites, suggesting that other factors control the dominant methane production pathway. Subsequent molecular biological studies confirmed that aerobic methanotrophic bacteria were also important and that their relative abundance was highest at the peat extraction site. Furthermore, the composition of the methanotrophic community varied between sites and depths. The aerobic methane oxidation rates were highest at the peat extraction site, potentially oxidizing a multiple of the emitted methane and, thus, likely providing an effective microbial methane filter. For the assessment of potential leakage from cut and buried abandoned wells, our results highlight the need to combine methane emissions with soil gas characterization in comparison to a suitable reference site. Monitoring that relies exclusively on methane emissions may result in the erroneous classification of naturally occurring emissions as well integrity failure.