@article{DeLuciaKuehnLindemannetal.2021, author = {De Lucia, Marco and K{\"u}hn, Michael and Lindemann, Alexander and L{\"u}bke, Max and Schnor, Bettina}, title = {POET (v0.1): speedup of many-core parallel reactive transport simulations with fast DHT lookups}, series = {Geoscientific model development : an interactive open access journal of the European Geosciences Union}, volume = {14}, journal = {Geoscientific model development : an interactive open access journal of the European Geosciences Union}, number = {12}, publisher = {Copernicus}, address = {G{\"o}ttingen}, issn = {1991-959X}, doi = {10.5194/gmd-14-7391-2021}, pages = {7391 -- 7409}, year = {2021}, abstract = {Coupled reactive transport simulations are extremely demanding in terms of required computational power, which hampers their application and leads to coarsened and oversimplified domains. The chemical sub-process represents the major bottleneck: its acceleration is an urgent challenge which gathers increasing interdisciplinary interest along with pressing requirements for subsurface utilization such as spent nuclear fuel storage, geothermal energy and CO2 storage. In this context we developed POET (POtsdam rEactive Transport), a research parallel reactive transport simulator integrating algorithmic improvements which decisively speed up coupled simulations. In particular, POET is designed with a master/worker architecture, which ensures computational efficiency in both multicore and cluster compute environments. POET does not rely on contiguous grid partitions for the parallelization of chemistry but forms work packages composed of grid cells distant from each other. Such scattering prevents particularly expensive geochemical simulations, usually concentrated in the vicinity of a reactive front, from generating load imbalance between the available CPUs (central processing units), as is often the case with classical partitions. Furthermore, POET leverages an original implementation of the distributed hash table (DHT) mechanism to cache the results of geochemical simulations for further reuse in subsequent time steps during the coupled simulation. The caching is hence particularly advantageous for initially chemically homogeneous simulations and for smooth reaction fronts. We tune the rounding employed in the DHT on a 2D benchmark to validate the caching approach, and we evaluate the performance gain of POET's master/worker architecture and the DHT speedup on a 3D benchmark comprising around 650 000 grid elements. The runtime for 200 coupling iterations, corresponding to 960 simulation days, reduced from about 24 h on 11 workers to 29 min on 719 workers. Activating the DHT reduces the runtime further to 2 h and 8 min respectively. Only with these kinds of reduced hardware requirements and computational costs is it possible to realistically perform the longterm complex reactive transport simulations, as well as perform the uncertainty analyses required by pressing societal challenges connected with subsurface utilization.}, language = {en} } @article{TranterDeLuciaKuehn2021, author = {Tranter, Morgan Alan and De Lucia, Marco and K{\"u}hn, Michael}, title = {Numerical investigation of barite scaling kinetics in fractures}, series = {Geothermics : an international journal of geothermal research and its applications}, volume = {91}, journal = {Geothermics : an international journal of geothermal research and its applications}, publisher = {Elsevier}, address = {Amsterdam [u.a.]}, issn = {0375-6505}, pages = {14}, year = {2021}, abstract = {Barite stands out as one of the most ubiquitous scaling agents in deep geothermal systems, responsible for irreversible efficiency loss. Due to complex parameter interplay, it is imperative to utilise numerical simulations to investigate temporal and spatial precipitation effects. A one-dimensional reactive transport model is set up with heterogeneous nucleation and crystal growth kinetics. In line with geothermal systems in the North German Basin, the following parameters are considered in a sensitivity analysis: temperature (25 to 150 degrees C), pore pressure (10 to 50 MPa), fracture aperture (10(-4) to 10(-2) m), flow velocity (10(-3) to 10(0) m s(-1)), molar volume (50.3 to 55.6 cm(3) mol(-1)), contact angle for heterogeneous nucleation (0 degrees to 180 degrees), interfacial tension (0.07 to 0.134 J m(-2)), salinity (0.1 to 1.5 mol kgw(-1) NaCl), pH (5 to 7), and supersaturation ratio (1 to 30). Nucleation and consequently crystal growth can only begin if the threshold supersaturation is exceeded, therefore contact angle and interfacial tension are the most sensitive in terms of precipitation kinetics. If nucleation has occurred, crystal growth becomes the dominant process, which is mainly controlled by fracture aperture. Results show that fracture sealing takes place within months (median 33 days) and the affected range can be on the order of tens of metres (median 10 m). The presented models suggest that barite scaling must be recognised as a serious threat if the supersaturation threshold is exceeded, in which case, large fracture apertures could help to minimise kinetic rates. The models further are of use for adjusting the fluid injection temperature.}, language = {en} } @article{DeLuciaKuehn2021, author = {De Lucia, Marco and K{\"u}hn, Michael}, title = {DecTree v1.0-chemistry speedup in reactive transport simulations}, series = {Geoscientific model development : an interactive open access journal of the European Geosciences Union}, volume = {14}, journal = {Geoscientific model development : an interactive open access journal of the European Geosciences Union}, number = {7}, publisher = {Copernicus}, address = {G{\"o}ttingen}, issn = {1991-959X}, doi = {10.5194/gmd-14-4713-2021}, pages = {4713 -- 4730}, year = {2021}, abstract = {The computational costs associated with coupled reactive transport simulations are mostly due to the chemical subsystem: replacing it with a pre-trained statistical surrogate is a promising strategy to achieve decisive speedups at the price of small accuracy losses and thus to extend the scale of problems which can be handled. We introduce a hierarchical coupling scheme in which "full-physics" equation-based geochemical simulations are partially replaced by surrogates. Errors in mass balance resulting from multivariate surrogate predictions effectively assess the accuracy of multivariate regressions at runtime: inaccurate surrogate predictions are rejected and the more expensive equation-based simulations are run instead. Gradient boosting regressors such as XGBoost, not requiring data standardization and being able to handle Tweedie distributions, proved to be a suitable emulator. Finally, we devise a surrogate approach based on geochemical knowledge, which overcomes the issue of robustness when encountering previously unseen data and which can serve as a basis for further development of hybrid physics-AI modelling.}, language = {en} } @article{TranterDeLuciaKuehn2021, author = {Tranter, Morgan Alan and De Lucia, Marco and K{\"u}hn, Michael}, title = {Barite scaling potential modelled for fractured-porous geothermal reservoirs}, series = {Minerals}, volume = {11}, journal = {Minerals}, number = {11}, publisher = {MDPI}, address = {Basel}, issn = {2075-163X}, doi = {10.3390/min11111198}, pages = {22}, year = {2021}, abstract = {Barite scalings are a common cause of permanent formation damage to deep geothermal reservoirs. Well injectivity can be impaired because the ooling of saline fluids reduces the solubility of barite, and the continuous re-injection of supersaturated fluids forces barite to precipitate in the host rock. Stimulated reservoirs in the Upper Rhine Graben often have multiple relevant flow paths in the porous matrix and fracture zones, sometimes spanning multiple stratigraphical units to achieve the economically necessary injectivity. While the influence of barite scaling on injectivity has been investigated for purely porous media, the role of fractures within reservoirs consisting of both fractured and porous sections is still not well understood. Here, we present hydro-chemical simulations of a dual-layer geothermal reservoir to study the long-term impact of barite scale formation on well injectivity. Our results show that, compared to purely porous reservoirs, fractured porous reservoirs have a significantly reduced scaling risk by up to 50\%, depending on the flow rate ratio of fractures. Injectivity loss is doubled, however, if the amount of active fractures is increased by one order of magnitude, while the mean fracture aperture is decreased, provided the fractured aquifer dictates the injection rate. We conclude that fractured, and especially hydraulically stimulated, reservoirs are generally less affected by barite scaling and that large, but few, fractures are favourable. We present a scaling score for fractured-porous reservoirs, which is composed of easily derivable quantities such as the radial equilibrium length and precipitation potential. This score is suggested for use approximating the scaling potential and its impact on injectivity of a fractured-porous reservoir for geothermal exploitation.}, language = {en} } @article{TranterDeLuciaWolfgrammetal.2020, author = {Tranter, Morgan Alan and De Lucia, Marco and Wolfgramm, Markus and K{\"u}hn, Michael}, title = {Barite scale formation and injectivity loss models for geothermal systems}, series = {Water}, volume = {12}, journal = {Water}, number = {11}, publisher = {MDPI}, address = {Basel}, issn = {2073-4441}, doi = {10.3390/w12113078}, pages = {24}, year = {2020}, abstract = {Barite scales in geothermal installations are a highly unwanted effect of circulating deep saline fluids. They build up in the reservoir if supersaturated fluids are re-injected, leading to irreversible loss of injectivity. A model is presented for calculating the total expected barite precipitation. To determine the related injectivity decline over time, the spatial precipitation distribution in the subsurface near the injection well is assessed by modelling barite growth kinetics in a radially diverging Darcy flow domain. Flow and reservoir properties as well as fluid chemistry are chosen to represent reservoirs subject to geothermal exploration located in the North German Basin (NGB) and the Upper Rhine Graben (URG) in Germany. Fluids encountered at similar depths are hotter in the URG, while they are more saline in the NGB. The associated scaling amount normalised to flow rate is similar for both regions. The predicted injectivity decline after 10 years, on the other hand, is far greater for the NGB (64\%) compared to the URG (24\%), due to the temperature- and salinity-dependent precipitation rate. The systems in the NGB are at higher risk. Finally, a lightweight score is developed for approximating the injectivity loss using the Damkohler number, flow rate and total barite scaling potential. This formula can be easily applied to geothermal installations without running complex reactive transport simulations.}, language = {en} }