A Geothermal Well Doublet for Research and Heat Supply of the TU Delft Campus
Philip Vardon, David Bruhn, Abe Steiginga, Barbara Cox, Hemmo Abels, Auke Barnhoorn, Guy Drijkoningen, Evert Slob, Kees Wapenaar
AA Geothermal Well Doublet for Researchand Heat Supply of the TU Delft Campus
Philip J. Vardon, David F. Bruhn, Abe Steiginga, Barbara Cox, Hemmo Abels, AukeBarnhoorn, Guy Drijkoningen, Evert Slob, Kees Wapenaar
Summary a r X i v : . [ phy s i c s . g e o - ph ] M a r . INTRODUCTION Though the number of geothermal well doublets in operation has been increasing overthe last years, none of these wells has been equipped with more than the bare minimum ofmonitoring equipment, and for none of them were detailed geological and geophysical baseline data collected. Despite the success of implementing this technology at a reasonablescale, many questions related to practical and more fundamental operational aspects aretherefore still open. These include classical issues such as the uncertainty about productivityand temperature of the resource prior to drilling, but also problems related to downholeprocesses and fluid chemistry and operational decision making. These fundamental issuesoften lead to operational problems such as poor injectivity, corrosion of installations andprecipitation of solids. In addition, our limited understanding of the thermal, hydraulic,mechanical and chemical processes in the subsurface related to geothermal operations resultin sub-optimal exploitation of the resource and potentially in technically and/or economicallyfailed projects. Moreover, the link to heat demand and heat distribution in a heterogeneoussystem (such as an urban environment) has not been extensively studied.Worldwide, research on deep geothermal wells has been focused on high-enthalpy heatproduction (water > < II. RESEARCH PROGRAMME
The research programme has received funding to provide the research infrastructure viathe EPOS-NL project. The research activities associated with this are made up from a seriesof ongoing and future projects, such that an overall programme can evolve over the lifetimeof the well to answer contemporary salient questions.
A. EPOS-NL • A Groningen gas field seismological network and data centre (KNMI). • The Earth Simulation Laboratory for multi-scale, rock physics and analogue experi-ments (UU). • A distributed facility for multi-scale imaging and tomography (MINT) of geo-materials. • The campus deep geothermal research facility (TUD).Incorporation of EPOS-NL within the European EPOS research infrastructure facilitatestransnational access to physical facilities, as well as optimal exploitation of research resultsvia open access data services. 3 . Main questions / ambition
With the DAPwell research programme we will create a world-class geothermal energyresearch facility where the following key topics will be addressed:1. Predictive power of flow models for optimal control and monitoring.2. Hydraulic-Thermal-Chemical behaviour: • Chemistry of geothermal fluids and their interaction with reservoir rocks andtechnical installations. • Monitoring of travelling fluid and cold fronts. • Advanced inflow estimation.3. Effect of human activities in the subsurface on the natural and built environment atsurface.4. Novel well completion: composite casing material.5. Integrated domestic heat management.6. Site specific characterisation: geological history, heterogeneities, reservoir fluids (wa-ter/gas/oil).Research undertaken in conjunction with the DAPwell will result in knowledge for furtherinnovation in geothermal science and engineering on topics that, at the moment, hamperthe ability to fully exploit the potential of this technology.The research will target a better prediction of the lifetime of a geothermal doublet, i.e.when the breakthrough of the cold front at the production well occurs, meaning that heatproduction substantially reduces. The approach suggested is (a) geophysical monitoringusing the full range of active-source electromagnetics, passive electromagnetics and seismics,chemical tracers, cross-well pressure transient analysis, with (b) data-assimilation methodsto integrate the results into prediction tools. In addition, the system integrity will beimproved, with an innovative approach to prevent corrosion. The approach proposed isto (a) investigate none-corrosive casing materials and study their long-term performance,(b) understanding the fluid chemistry as a function of pressure and temperature, and (c)4eveloping modelling tools to predict the chemical interaction of the complex fluid withrocks and casing. Moreover, the wider energy system will be investigated. The designeddoublet will exceed the heat demand of the campus and provide heat when there is nodemand, therefore the management and extension of the heat grid to outside the campuswill be investigated.In addition, we anticipate that new scientific questions and need for innovations will occurduring the lifetime of this large-scale long-term project. Therefore, flexibility is built intothe scientific design.
III. PROJECT DESIGN /hr and produce hotwater at approximately 75C. Current planning foresees that the return temperature to thesystem will be initially around 50C, although it is anticipated that this will reduce overtime, as buildings are renovated and further buildings are included in the heat distributionsystem. TU Delft will withdraw at least 100GJ of heat energy per year to provide heating tothe campus. To enable an easy connection to the existing heat grid, the surface location ofthe wells is planned to be close to the existing heat grid in a location where sufficient space isavailable for the required construction infrastructure, indicated approximately in Figure 1(a)by the open blue square on the left hand side. The wells will follow the schematic trajectory5 round surface~800m depth ~2200m depth~1400moutstep (a) (b) FIG. 1: (a) The designed geothermal location, with the reservoir top and bottom indicated bycircles (red for producer and blue for injector). The black boxes are the license areas, includingtwo adjacent projects and the green triangles are the proposed surface geophysical shallow boreholearray locations. (b) Designed well trajectory, note that the outsteps are not on the same plane. shown in Figure 1(b), where vertical sections from the ground surface to approximately 800m depth followed by deviated sections at approximately 45 until the reservoir. This meansthat each well will outstep by approximately 1400 m, both towards where the reservoir isdeeper (i.e. not in the opposite direction from each other), as indicated in Figure 1(a) bythe red and blue circles, indicating the well intersection with the top and bottom of thereservoir. A 3D representation of the reservoir is shown in Figure 2.The diameter of the production well is planned to be approximately 8 inch in thereservoir section, with a glass reinforced epoxy (GRE) lined steel tubing foreseen above thereservoir in the lower straight (deviated) section, increasing in size until the surface. Anexternal casing is planned, which will result in an annulus between the production tubing andexternal casing. This annulus will be used for (i) monitoring to detect the performance ofthe casing and production tubing and (ii) as a carrier for the fibre optic monitoring system.This means that the fibre optic could be replaced if needed or if new instrumentation isproposed. A submersible pump is designed to be inserted at approximately 850 m depthwith a production tubing to surface. The injection well is proposed to be approximately9 inch in the reservoir with a liner running to surface. Larger sections (liners) will beconstructed closed to surface. This well is also proposed to be GRE lined from above thereservoir to the surface (WEP, 2019). 6 ource: Douglas Gilding FIG. 2: 3D view of the reservoir including the planned wells and existing geothermal projects andexploratory wells.
The geology derived from seismic lines and exploration wells nearby, is shown in Figure3. The proposed coring plans are shown alongside. Due to the interest in oil and gasexploration, there is less known in the region between approximately 200 1700 m depth. Inparticular there is known to be a series of unconformities at around 400 m deep. The upperlayers are also of interest for heat storage. The complete reservoir, along with the cap rockand underburden are planned to be cored.
IV. LOGGING AND MONITORING PROGRAMME
One of the main objectives of the research is to fully understand the production andbehaviour of the flow in the reservoir. To do this well logging both prior to casing andafterwards and ongoing monitoring is planned in both wells.The proposed logging programme in the open holes will include • SP and natural gamma-ray (plus spectral gamma-ray) • Resistivity – deep, shallow and micro laterolog (LLD, LLS, microL)7 roup Period Formation (important members) Depth(mTVD) CoringUpper North Sea Quaternary Maassluis 0Tertiary Oosterhout 300 ~200mBreda 395
Middle North Sea
Rupel 400
Lower North Sea
Landen 405
Chalk
Upper Cretaceous Ommelanden 415Texel 440
Rijnland
Lower Cretaceous Holland 520 ~30m@950mVlielandSandstone 1200 ~30m@1450m~30m@1650m
Schieland
Nieuwerkerk(Rodenrijs Claystone, Delft Sandstone, Alblasserdam) 1700 to 2000(injector and producer respectively) ~300m @ 2040 in producer (caprock, reservoir and underburden)
FIG. 3: Simplified lithography (after WEP, 2014) and coring plans. – Induction log - IL – micro spherically focussed log MSFL – dual-spacing neutron log (CNL) with compensated formation density log (FDC) • Sonic ( v p , v s , full waveforms) • Gamma Density • Nuclear magnetic resonance logging NMR (maybe, for permeability) • Micro dip meter tool, dip measurements • Fullbore micro-imager logs - FMI • Mini-frac tests at certain levels • Caliper logInside the casing: • Gamma ray • Induction logs • FDC / CNL 8
Sonic • Ultrasonic Imaging Tool (USIT) and casing collar locator (CCL) • Nuclear magnetic resonance logging NMR (maybe) • A cement bond log (CBL) to test the cementation of the novel casing, and an accuratecaliperThe programme may still have to be adapted based on risk and objectives. Some of theabove tools can be combined in one toolstring to reduce the openhole time. Fluid sampleswill be taken to perform pressure, volume, temperature (PVT) analysis of the reservoirfluids, after the first production test. Samples will have to be taken by means of wirelinewith memory gauge or E-line. For this the ESP will have to be removed after the productiontest.Standard production logging will take place, alongside a bespoke monitoring system. Themonitoring system is mainly based upon fibre optic technology. In the injection well, thefibre optic cable (a bundle of fibre optics) will be installed behind the inner liner, cementedin place to ensure a solid contact with the formation. This cable is designed to measuredistributed temperatures and acoustics. The acoustic sensors are planned to be wound in away such that multidirectional acoustics can be detected. In the production well, the fibreoptic cable will be attached to the outside of the production casing, which finishes above thereservoir, and will be supported inside the well through the reservoir. This cable will enabledistributed temperature and acoustic sensing (DTS and DAS), and additionally will includedistributed pressure sensing within the reservoir section. The pressure sensors will allow (i)temporal pressure analysis and (ii) differential flow measurement. These measurements willbe augmented by occasional downhole flow monitoring and local tracer measurements.To monitor the impact at the surface and downhole two vertical fibre optic cables willbe installed, and a dense seismic and electromagnetic monitoring network will be set up,providing a new basis for observation of processes and changes occurring following the op-eration of deep geothermal heating doublets. The design is intended to detect any inducedseismicity and allow the source to be spatially identified. Temperature and induced flow(convection) monitoring is planned in the shallow subsurface. Innovative methods of seismicinterferometry and virtual seismology under continuous development (e.g., Boullenger et9l. 2015, Wapenaar et al. 2018) will be applied to determine the location, source mecha-nism and radiation characteristics of seismicity to image and monitor cold-water injection inDAPWELL. The proposed locations for the geophysical array are shown by green trianglesin Figure 1(a).The borehole seismic measurements using the fibre-optic cables will be carried out in time-lapse mode using an active source, and in continuous monitoring mode using the acousticnoise generated by the water production/injection. As active seismic source, a highly in-novative seismic vibrator based on linear synchronous motors will be used that has beendeveloped at TU Delft (Fig. 4a). This source will be placed at the surface and allows gener-ation of low-distortion, very repeatable and wide-frequency-band signals, including the lowfrequencies.
FIG. 4: a) Innovative seismic vibrator based on linear synchronous motors that generates lowfrequencies (typically down to 1.5 or 2 Hz); b) Sketch of the electromagnetic network.
With active-source electromagnetics, high-resistivity bodies can be detected, localized,and monitored from the surface and with sensors in the wells (Wirianto et al., 2011, Schalleret al., 2019). Producing large volumes of hot and injecting cold water creates acoustic andelectromagnetic noise, which can be used as signal in so-called passive (ambient-noise) mea-surements. Using advanced signal processing methodology developed at TU Delft, these10assive measurements can be used in addition to the active-source electromagnetic measure-ments to improve the imaging and monitoring around the well. The major challenges arethe detection of a body that is relatively small compared to its depth and the influence ofanthropogenic noise in an urban environment.For the active electromagnetic (EM) monitoring a transient electric-dipole source willbe used at the surface, while electric and magnetic sensors will be placed on the surface(Fig. 4b) and in the boreholes. Pseudo random binary sequences (PRBS) will be used asthe source time-signal designed to improve the signal-to-noise ratio. This equipment canbe bought with possible adaptations to the source time signature. Downhole electrodesfor electric field measurements will be developed and placed for passive monitoring usingelectromagnetic noise generated by the water production/injection.The development of electrodes and housing of receiver electronics to endure permanentexposure to the ambient high pressure and temperature and the aggressive chemical com-position of the pore fluids is an outstanding technical challenge. No sensors seem to existat the moment that can operate continuously for long periods under these conditions. Incase the development of sensors that stand these conditions fails, alternative borehole sen-sors will be made that can be operated in time-lapse mode and will be removed from theborehole after the transient measurements. Electrodes are mostly sensitive to corrosion andthe challenge is to find a solution that provides resistance to corrosion while maintaininglow contact resistance over time. Receiver electronics housing will be designed and built inhouse with existing expertise. The most crucial development concerns the realisation of astable electrical contact between the electrodes and the terminals of the housing.
V. LABORATORY INFRASTRUCTURE
To make sure the wells can serve as reference wells for research, an extensive coring pro-gramme will be implemented during drilling. The cores will be stored and made availablefor laboratory testing, general characterisation and calibration of subsurface measurements.The petrophysical properties of the reservoir rocks will be determined in the PetrophysicsLaboratory at TU Delft, such as density, porosity and permeability, electrical and thermalconductivities as well as mechanical properties such as elastic moduli. The detailed miner-alogical and chemical analysis (X-ray methods, microscopy, SEM and electron microprobe,11n EPOS-NL MINT) will provide the basis for the evaluation of fluid-rock interaction andfor the monitoring and prediction of the long-term behaviour of the system.
VI. CONCLUSIONS
The DAPWELL research and monitoring infrastructure will be used to investigate thefundamental scientific challenges that are presently limiting the development of geothermalenergy. This could not be realized by using an existing geothermal plant that is built forcommercial energy production because the monitoring equipment cannot be installed onceoperation has started and drill cores are not taken in commercial wells. This exceptionaldevelopment will result in innovations that leapfrog from well-functioning doublets usedtoday to highly efficient geothermal installation in ten years time.Access to the high-level research infrastructure will render DAPWELL in particular andEPOS-NL in general not only a key national infrastructure but also result in major appealto talented researchers from elsewhere in Europe and beyond. Likewise, the universitypartners participate in a wide range of European programmes that will assure visibility,which is prerequisite for high-potential researchers to become interested in EPOS-NL andto be attracted by geothermal research.The unique aspect of the DAPWELL facility will be the possibility to do research using anoperating geothermal plant. It will be used as a laboratory where researchers study optimalproduction scenarios and monitoring techniques in order to achieve the highest possibleenergy efficiency. This could not be realized by using an existing geothermal plant that isbuilt for commercial energy production. The unique approach of research at DAPWELL willresult in innovations to improve the efficiency of geothermal systems. As there is currentlyno such research infrastructure in an operating geothermal well doublet worldwide, theDAPWELL will provide a unique opportunity for many high-potential researchers to doresearch with this fully operational research well. The uniqueness of the infrastructure willlead to productive cooperation, especially with European partner institutions (for example,the EERA joint programme) and the growing geothermal industry in the Netherlands andbeyond. 12
CKNOWLEDGEMENTS
We acknowledge funding from The Netherlands Organization for Scientific Research(NWO), National Roadmap Programme “EPOS-NL: The Netherlands contribution to theEuropean Plate Observing System”. In addition, the research of K. Wapenaar has receivedfunding from the European Research Council (ERC) under the European Unions Horizon2020 research and innovation program (grant no. 742703).