Induced geochemical reactions by compressed air energy storage in a porous formation in the North German Basin

Abstract

A synthetic diabatic compressed air energy storage (CAES) based on an existing surface facility and using the Rhaetian sandstone formation in the North German Basin as porous storage reservoir is investigated for induced geochemical reactions. A daily storage cycle with a 6-h injection and a 6-h extraction phase in the early morning and late afternoon is applied. The chosen mineral assemblage shows the presence of pyrite in the Rhaetian sandstone, so that operating CAES in this porous formation introduces oxygen in the formation and induces geochemical reactions governed mainly by pyrite oxidation. A consistent geochemical reaction system for the Rhaetian sandstone is developed and kinetic batch simulations are used to quantify the changes due to the induced geochemical reactions in both stored air as well as storage formation.

The injection of air into this geological formation leads to pyrite oxidation, changes in stored air composition, air pressure and formation properties. Results show that only a very small change up to 0.29% in the oxygen mole fraction is found within one storage cycle, which does not affect flammability of the stored air. Considering a longer residence time, the oxygen concentration in the stored air may drop below the minimum oxygen concentration for flame propagation and thus cannot be used for burning natural gas as required for a diabatic CAES. The pH of the formation fluid after 20-year cyclic daily operation can drop significantly below one near the gas wells increasing the risk of well corrosion, but with smaller effects at larger distances. However, mineral dissolution and precipitation found in the storage formation results in only minor increases of porosity and permeability with relative changes up to 1.0% and 5.0%, respectively. The uncertainties in mineral reactive surface area and pyrite oxidation kinetic strongly affect the rate of oxygen reduction and fluid acidification. Analysis of these parameters of the reservoir mineral phases from the target location, especially for pyrite, are therefore required for a reliable estimate of possible induced geochemical reactions and impacts.

Introduction

As a means of reducing greenhouse gas emission and mitigating climate change effects, renewable energy sources are increasingly substituting fossil fuels. Wind farms and solar power stations are the major sources for renewable energy production, which implies strong fluctuations in electric power generation in the short term due to changing weather conditions and in the longer term due to seasonality effects. To compensate for these fluctuations, a large storage demand of electrical energy is required, with estimates up to 50 TWh for Germany (Bräutigam et al., 2017) or 600 GWh for Denmark aiming for 80% renewable energy share (Sorknæs et al., 2013). Grid-scale standby storage systems in the geological subsurface provide promising storage options, due to their large potential storage capacities, high achievable input and output rates and their flexibility to be employed on time scales varying from hourly to seasonally (Kabuth et al., 2017).

Of these options, large-scale compressed air energy storage (CAES) in the subsurface is one of the gas storage options which is able to compensate strong fluctuations on the hourly to daily basis (Budt et al., 2016). CAES represents a “power to power” energy storage option (Sternberg and Bardow, 2015), which converts off-peak electricity to mechanical energy in the form of pressurized air, stores the compressed air in the subsurface geological formations and retrieves the stored air during times of peak electricity demand to produce power using gas turbines. Currently, only two CAES facilities, i.e. in Huntorf, Germany, and in McIntosh, US, are operating, both using subsurface salt caverns to store the compressed air (Kushnir et al., 2012). Because suitable geological salt formations for mining the required caverns are not widely available but porous formations are more frequently occurring, porous formation storage of compressed air becomes a promising world-wide option (Succar and Williams, 2008). Moreover, porous formations can provide much larger potential storage capacities (Kabuth et al., 2017). Although currently no CAES in porous formations is operating, the concept has been shown feasible in first studies in the 1980s at a field test site in Pittsfield, Illinois, USA (ANR Storage Company, 1990).

The operation of CAES using porous formations as storage reservoirs is governed by thermal (T), hydraulic (H), mechanical (M) and chemical (C) processes and induces relevant impacts, and a reliable quantification of these is required for designing and dimensioning a storage operation (Bauer et al., 2015, 2013). In terms of the induced hydraulic, mechanic and thermal processes, CAES is comparable to natural gas storage or carbon dioxide capture and storage, which are well known from decades of experience in the oil and gas industry as well as recent research efforts. Anticlinal sites show especially high potential (e.g. Mitiku and Bauer (2013)) for CAES in a porous formation, therefore, the induced pressure is highest as well as lowest at the gas wells and needs to stay within the pressure thresholds for formation damage (Wang and Bauer, 2017a). Due to the expected short operational cycles in CAES, the pressure response during the cyclic operation will be restricted mainly to the gas reservoir (Wang and Bauer, 2017b). The induced temperature change in the storage formation can be limited to only a few Kelvin, as the temperature of injected air from the compressor is cooled down close to the reservoir temperature (Oldenburg and Pan, 2013a). A cyclic operation may introduce cyclic stresses in the formation rock, and Erikson (1983) found a loss of 22% in permeability during the first cycle due to the hysteresis effect for St. Peter sandstone, and much smaller decreases in subsequent cycles as the sandstone shows an elastic behavior.

However, the potentially induced chemical impacts by CAES in a porous formation differ significantly from those of other gas storages, as air containing oxygen is introduced into porous geological formations long-free of oxygen. If redox-sensitive or ferrous-containing minerals are present, such as pyrite (FeS₂, iron disulfide), oxidation processes will be induced. These oxidation processes are known from carbon dioxide storage considering oxygen as an impurity (André et al., 2015; Jung et al., 2013; Pearce et al., 2016b, 2016a; Wei et al., 2015), but not studied for the case of CAES applications. Operating CAES in a pyrite containing formations induces pyrite oxidation, and thus the oxygen in the stored air will be partly or completely consumed (ANR Storage Company, 1990). Without a refill of the storage reservoir, the reduced fraction of oxygen in the stored air may potentially cause a failure of a diabatic CAES facility, because the air extracted from the storage may not contain enough oxygen for the gas combustion process required to heat the expanding gas. Studies on acid mine drainage indicate that pyrite oxidation with on-going supply of oxygen, e.g. near gas wells for CAES operation, can lower the pH to very acidic conditions (INAP, 2012; Nordstrom et al., 2015), which thus increases the risk of wellbore corrosion at the gas storage wells. Meanwhile, mineral precipitation induced by geochemical reactions may clog the pore space, thus reducing porosity and permeability of the storage formation, which would again lower the well deliverability and power output (Pei et al., 2015).

Due to these induced geochemical reactions and potential impacts, a reliable quantification of induced geochemical reactions is a prerequisite for assessing the feasibility of porous media CAES. This study therefore focusses on investigating and quantifying potential induced geochemical impacts of a porous formation CAES operation. Potential porous media storage formations are present in the North German Basin (Hese, 2012, 2011), which stretches over parts of Germany, Poland, Denmark, the Netherlands and Great Britain. In this work, the Rhaetian sandstone formation is investigated as a potential storage formation for CAES, as it is well characterized and has been shown suitable in previous research for use in hydrogen storage (Pfeiffer et al., 2017, 2016a; Pfeiffer and Bauer, 2015) or CO₂ storage (Mitiku et al., 2013). As the mineral composition of the Rhaetian sandstone shows the presence of pyrite (Dethlefsen et al., 2014; Mitiku et al., 2013), the induced pyrite oxidation and other geochemical reactions are investigated under a consistent geochemical reaction system for the Rhaetian sandstone. The induced impacts on the stored air as well as the storage formation are also quantified using scenario analysis and process based kinetic batch modelling of such a geochemical system.