Preventing loss of near-well permeability in CO2 injection wells

CO2 storage projects involve the pumping CO2 from injection wells into porous underground rock formations. An important prerequisite is that the injection can be done in a safe and efficient manner, and this requires sufficient flow of CO2 from the well into the formation. CO2 injection has shown to be challenging in several cases, since there are many potential mechanisms than can lead to clogging of the near wellbore area. In this project, we have chosen to study some of the clogging mechanisms from an operational perspective; that is, we will try to understand how the conditions in the near wellbore area, such as stresses, pressure, temperature, flow, and rock type, as well as stops and starts in CO2 injection, influence the risk of clogging. This could be caused by e.g. salt precipitation, wax or hydrate formation, bacterial films, clay swelling, and transport of fines from partially crushed rock around the well. We have focused on the problem of salt precipitation, which is caused by the brine situated in the CO2 storage being dried by the dry dry CO2 injected from the well into the porous rock. Continuous drying of the brine can lead to the nucleation and growth of salt crystals that clog the pore space and prevents further injection of CO2. Exactly how this happens and if it becomes a problem has not been well understood. We have investigated this issue on several length scales. We have developed an equation of state for predicting under which circumstances salt precipitation is realistic. This is implemented in the open source code MRST. We then simulated how salt crystals are created and grow in the pore space, and how this reduces the flow capacity (permeability) of a porous medium such as a reservoir rock. We have modelled injection of CO2 into rock to predict how the rock dries and salt is deposited. Further, we have performed a series of small scale injection experiments into different rock types to investigate how different parameters such as rock type, injection rate, flow direction, and brine salinity affects salt precipitation and clogging. Finally, we designed an injection experiment in a downscaled well geometry to observe these mechanisms in a more realistic situation.

For det første har prosjektet ført til samarbeid på tvers av flere institusjoner i et fagområde som er komplekst, og hver forskningspartner har bidratt med med sin kjernekompetanse for å understøtte prisosjektets mål. Grunnleggende og sterk kompetanse på feltet er bygget ved å kunne arbeide over tid på den relativt avgrensede, men komplekse problemstillingen saltutfelling. Det faglige arbeidet utviklet ny kunnskap på områder som tidligere var mangelfulle, slik som (1) modeller for hvordan salt feller ut og vokser i porerommet, og effekten av dette på gjennomstrømningsevne nær brønnen, (2) hvordan saltutfelling og tetting av stein avhenger av steintype, strømningsretning, o.l., og (3) hvordan salt feller ut i mer realistiske brønngeometrier. Arbeidet med nye tilstandsligninger er implementert i åpen kode (MRST), og er tatt i bruk i lignende problemstillinger i f.eks. hydrogenlagring. En PhD-kandidat ble utdannet og arbeider innen dette feltet nå. Kunnskapen og anbefalingene fra prosjektarbeidet kan føre til en bedre forståelse for hvorvidt saltutfelling blir et problem for operatører under injeksjon av CO2 i saltvannførende lag, som er de mest aktuelle lagrene på norsk sokkel. Det er tydeligere hva slags eksperimenter man kan utføre for å få kunnskap om risikoen for tetting i et spesifikt reservoar. Det kan bidra i planleggingen av komplettering, eller hjelpe å forutsi problemer eller gi føringer for hvordan man bør injisere.

This project aims to study how a high permeability in the near-well region can be maintained during intermittent CO2 injection and backflow of CO2-brine. Linear- and core-scale experiments will be performed to pinpoint the most relevant injectivity loss mechanisms at realistic field conditions, and unique radial flow experiments will thereafter be performed in a true triaxial test cell. This enables studies of not only precipitation reactions in pores (e.g. salt/wax formation), but also fines migration and loss of permeability due to grain displacement or shale swelling. Promising methods for injectivity loss mitigation (e.g. chemical injections) will also be tested. The experimental results will be used in a simulator to predict injectivity loss in the field, with a special test case related to the full-scale storage pilot at Smeaheia.