Live CO2 Storage Simulator — Vertical-Equilibrium Engine

SubsurfaceAI | 2D depth-integrated two-phase CO2/brine FD simulator (Nordbotten–Celia VE class) | AER D065 / EPA Class VI screening & AOR support
Physics envelope: depth-integrated vertical-equilibrium sharp-interface two-phase flow with Land (1968) residual trapping and pressure-dependent rate-limited dissolution (Duan-anchored equilibrium) — the model class used for CCS screening and AOR delineation (accepted in EPA Class VI AoR evaluations and AER D065 supporting analyses). Full governing equations and references are documented on the Theory & References tab. For the full 3D compositional containment demonstration, pair with the CMG-GEM decks (Wabamun_Nisku_CCS_V8*.dat) built alongside this tool.

Grid & Geometry

3D: NX ≤ 40 recommended (runtime ~10–60 s). 2D VE: NX 50+ fine.

Rock

3D Layer Stack (K1 top → KN base)

Kφkh (mD)
Used when Engine = 3D. Kv/Kh:

Initial Conditions & Fluids

Relative Permeability & Trapping

kd = first-order rate constant for approach to Duan-anchored equilibrium dissolved fraction (P-, T-, salinity-dependent). Default 1.2 /yr ≈ 0.6 yr half-time.

Schedule & Boundary

Injection Wells

Namex (m)y (m)Mt/yrstart yrstop yrBHPmax kPa
Schedule:
Blank stop yr = inject until end of injection period. Wells never inject past the injection period set in Schedule & Boundary.

Monitoring Wells

Namex (m)y (m)
idle
Ready. Configure the model, then Run.
VE reconstruction: CO2 column thickness h = Sg·H/(1−Swi) drawn from formation top (gravity override).
One-at-a-time sensitivity on a coarsened grid (NX 40). Metrics: plume radius at end of injection and max BHP. Select parameters and run.

Run the simulation first.

Data downloads (CSV — universal format)

AccuMap plume export (DXF / Shapefile)

Exports the Sg ≥ threshold plume outline for each snapshot year as geo-referenced closed polygons, plus injectors/monitors as points. Import into AccuMap via Import Manager: DXF converts to a user layer (points, lines, polygons); the zipped Shapefile works in AccuMap, geoSCOUT, Petra and ArcGIS. Coordinates are geographic lat/long — choose the datum matching your AccuMap session (Canada default is NAD27).

Set the anchor to your actual injector surface location so the plume lands on the correct township in AccuMap. Each year is a separate DXF layer (PLUME_YR_…) / shapefile record with YEAR and AREA_KM2 attributes, so you can display plume growth year by year.

Report

Generates a print-ready regulatory summary (parameters, results, compliance tables, figures). Use the browser print dialog → Save as PDF.

Audience: reservoir / CCS engineer reviewer (5-minute skim). This tab documents the governing equations, constitutive models, trapping mechanisms, numerical scheme, and analytical complements implemented in the simulator, with citations to the original sources. Every equation below is wired into the JavaScript engine — file references point to index.html line numbers.

0. Scope & model lineage

This tool is a depth-integrated vertical-equilibrium (VE) sharp-interface two-phase CO₂/brine simulator with an optional fully-resolved 3D layered engine. The VE class trades vertical resolution for speed (seconds per scenario) and is the standard screening-tier model used in CCS site characterization, AoR delineation, and Class VI permit-support work. Lineage: Coats et al. (1971) [1] introduced VE in steamflood modelling; Nordbotten & Celia (2005) [2] formalised it for CO₂ storage; the MRST-VE module (Lie 2019) [3] is the open-source reference implementation.

What it is for: screening, sensitivity analysis, AoR (40 CFR 146.84), AER D065 supporting analysis, multi-well rate optimisation. What it is NOT for: compositional containment demonstration (use the paired CMG-GEM V8 decks), geochemical mineral trapping over centuries, geomechanical fault reactivation, or wellbore integrity. Honest limits are listed in §6.

1. Governing equations

(1.1) Component mass conservation for phase $\alpha \in \{g, w\}$ over depth-integrated cell volume:

$$\frac{\partial}{\partial t}\!\left(\phi\, H\, \rho_\alpha\, S_\alpha\right) + \nabla \!\cdot\! \left(\rho_\alpha\, \mathbf{u}_\alpha\right) \;=\; q_\alpha \;\pm\; r_{diss}$$

Gas phase carries a dissolution sink $-r_{diss}$; brine phase carries the corresponding $+r_{diss}$. Code: index.html:428–529 (VE), :730–903 (3D).

(1.2) Multiphase Darcy:

$$\mathbf{u}_\alpha \;=\; -\frac{k\, k_{r\alpha}(S_g)}{\mu_\alpha}\bigl(\nabla P - \rho_\alpha g\, \nabla z\bigr)$$

Capillary pressure neglected at VE screening tier (typical CCS reservoirs at supercritical conditions: $P_c \ll \Delta P$). Code: pressure assembly :373–426 (VE), :661–728 (3D).

(1.3) Vertical-equilibrium reduction (Nordbotten–Celia 2005): assume vertical pressure gradient is hydrostatic in each phase and saturations stratify by gravity. CO₂ column thickness from the formation top:

$$h(x,y,t) \;=\; \frac{S_g(x,y,t)\, H}{1 - S_{wi}}$$

Used for cross-section reconstruction (:1374) and gravity-spreading flux (:485–500).

(1.4) VE gravity spreading flux (sharp-interface):

$$F_{grav} \;=\; \frac{k\, k_{rg,\max}}{\mu_g}\,(\rho_w - \rho_g)\, g\, h\, \frac{\partial h}{\partial x}$$

Direct implementation at :491–493; stability-capped at 20% of cell pore volume per substep.

2. Constitutive models

PropertyModel usedCodeReference
CO₂ density $\rho_g(P,T)$Bilinear interpolation of a Span–Wagner-anchored table (8–30 MPa, 30–80 °C):271–290Span & Wagner 1996 [4]
CO₂ viscosity $\mu_g(\rho)$$\mu_g = (0.015 + 2{\cdot}10^{-5}\rho + 5.2{\cdot}10^{-8}\rho^2)\times 10^{-3}$ Pa·s:292Fenghour–Wakeham–Vesovic 1998 [5] fit
Brine density $\rho_w(T, TDS)$$\rho_w = 1000 - 0.40(T-20) + 7{\cdot}10^{-4}\,TDS$ kg/m³:295Batzle & Wang 1992 [6] screening fit
Brine viscosity $\mu_w(T, TDS)$Vogel–Tammann–Fulcher water + salinity correction:296–299Kestin et al. 1981 [7]
Equilibrium dissolved CO₂ fraction $w^{eq}_{CO_2}$$0.050\,(P/12.3)^{0.55}\,[1-0.0045(T-55)]\,e^{-0.12 m}$ (kg/kg brine):301–306Duan & Sun 2003 [8] anchored
Rel-perm $k_{rg}, k_{rw}$VE: linear; 3D: Corey ($n_g{=}2$, $n_w{=}4$); $S_e = (S_g - S_{gr})/(1 - S_{wi} - S_{gr})$:363–368, :652–657Brooks & Corey 1964 [9]
Residual gas trapping $S_{gr}(S_{g,\max})$$S_{gr} = S_{g,\max}/(1 + C\, S_{g,\max})$, $C = 1/S_{gr,\max} - 1/(1-S_{wi})$:343, :357–362Land 1968 [10]

3. Trapping mechanisms (IPCC SRCCS classification)

#MechanismEquation / criterionThis simulator
1Structural / stratigraphicCaprock seal capacity vs. column buoyancy: $\Delta P_{buoy} = (\rho_w - \rho_g)\,g\,h_{col} < P_{e,seal}$; injection BHP $< 90\%\, P_{frac}$Implicit — closed-boundary option + fracture-pressure compliance check at :349, :1131–1134
2Residual (capillary) trappingLand (1968) hysteresis, eq. above; $S_{gr,\max}$ user input (default 0.12)Active — :343, :357–362, :536–538
3Solubility (dissolution) trapping$\dot m_{diss} \;=\; \alpha\,\dfrac{k_v\,\Delta\rho_c\,g}{\mu_w}\,C_{sat}\,A_{cell}$ — convection-limited (Rayleigh) flux at the gas–brine contact, $\alpha = 0.017$ (steady Sherwood), $C_{sat} = w^{eq}_{CO_2}\rho_w$, $\Delta\rho_c \approx 0.3\,C_{sat}$; capped by free-gas mass and full-column brine capacity; $w^{eq}$ re-evaluated each step at current cell pressureActive — :509–526 (VE), :875–898 (3D); $k_d$ knob scales $\alpha$
4Mineral trapping (carbonate precipitation)$\mathrm{CO_2(aq) + H_2O + Ca^{2+} \rightarrow CaCO_3 + 2H^+}$ on $10^2$–$10^4$ yr timescaleNot modeled — out of scope for ≤100 yr screening; use TOUGHREACT / GEM-GHG for kinetics

Framework: IPCC (2005) [11]; trapping inventory accounting follows Bachu (2015) [12]. Mass-balance partitioning of free / residual / dissolved is reported at :536–552 and rendered on the Time-Series tab.

4. Numerical scheme

StageMethodCode
PressureFully implicit IMPES with total-mobility weighting on cell faces (harmonic $k$, arithmetic $\lambda_T$); preconditioned conjugate gradient (Jacobi), tol $10^{-12}$ rel.:373–426, :661–728
TransportExplicit upwind fractional-flow $f_g = \lambda_g/(\lambda_g+\lambda_w)$ on cell-centered finite volumes:472–507, :813–873
CFL substepping$\Delta t_{sub} = \Delta t / n_{sub}$, with $n_{sub} = \lceil \max_c |\sum_n T_{cn}(P_c-P_n)|\,\Delta t / (0.25\, V_{p,c}) \rceil$:452–469
Gravity spreadingVE: sharp-interface buoyant flux capped at 20% of $V_p$ per substep. 3D: counter-current buoyancy with mobility-weighted face density:485–500, :827–846
Well modelPeaceman (1978) [13] well index $WI = 2\pi k H\,\mathrm{NTG}/\ln(0.2\,\Delta x/r_w)$; rate target with BHP cap; 3D engine distributes by layer-WI:436–448, :632–636, :753–770
Mass-balance closure$MBE\% = 100 (m_{free}+m_{trap}+m_{diss}-m_{inj})/m_{inj}$; compliance target $|MBE| < 1\%$:552, :931

5. Far-field pressure (Theis 1935)

Beyond the grid, single-phase brine pressure transient is approximated by the Theis solution to bracket the AoR pressure-front radius (criterion: $\Delta P > 10$ kPa per 40 CFR 146.84(c)(1)):

$$\Delta P(r,t) \;=\; \frac{Q\,\mu_w}{4\pi\, k\, H_{eff}}\, W(u), \qquad u \;=\; \frac{r^2\,\phi\,\mu_w\,c_t}{4\,k\,t}, \qquad W(u) \approx -\gamma - \ln u$$

Total compressibility $c_t = c_r + c_w \approx c_r + 4.6{\cdot}10^{-10}$ Pa⁻¹. Implemented at :575–588; results enter the regulatory AoR table.

6. Regulatory acceptance & honest limits

Accepted use: EPA Class VI 40 CFR 146.84 allows "computational modeling of appropriate complexity" for AoR delineation; VE/sharp-interface models are explicitly cited in EPA's AoR delineation guidance [14]. AER Directive 065 [15] accepts screening-class numerical simulation as supporting evidence in Scheme Applications.

Limits this model does not cover (use full compositional 3D + geomechanics for these):

7. References

  1. Coats, K.H., George, W.D., Chu, C., Marcum, B.E. (1971). Three-Dimensional Simulation of Steamflooding. SPE Journal 14(6), SPE-3416. — Origin of vertical-equilibrium reduction.
  2. Nordbotten, J.M., Celia, M.A. (2005). Similarity solutions for fluid injection into confined aquifers. J. Fluid Mech. 561, 307–327; and Water Resour. Res. 41, W12404. — VE theory for CO₂ storage.
  3. Lie, K.-A. (2019). An Introduction to Reservoir Simulation Using MATLAB/GNU Octave: User Guide for MRST. Cambridge Univ. Press. — Open-source VE reference implementation (MRST-co2lab).
  4. Span, R., Wagner, W. (1996). A New Equation of State for Carbon Dioxide Covering the Fluid Region from the Triple-Point Temperature to 1100 K at Pressures up to 800 MPa. J. Phys. Chem. Ref. Data 25(6), 1509–1596.
  5. Fenghour, A., Wakeham, W.A., Vesovic, V. (1998). The Viscosity of Carbon Dioxide. J. Phys. Chem. Ref. Data 27(1), 31–44.
  6. Batzle, M., Wang, Z. (1992). Seismic properties of pore fluids. Geophysics 57(11), 1396–1408.
  7. Kestin, J., Khalifa, H.E., Correia, R.J. (1981). Tables of the dynamic and kinematic viscosity of aqueous NaCl solutions in the temperature range 20–150 °C and the pressure range 0.1–35 MPa. J. Phys. Chem. Ref. Data 10(1), 71–87.
  8. Duan, Z., Sun, R. (2003). An improved model calculating CO₂ solubility in pure water and aqueous NaCl solutions from 273 to 533 K and from 0 to 2000 bar. Chem. Geol. 193, 257–271.
  9. Brooks, R.H., Corey, A.T. (1964). Hydraulic Properties of Porous Media. Hydrology Papers 3, Colorado State University.
  10. Land, C.S. (1968). Calculation of imbibition relative permeability for two- and three-phase flow from rock properties. SPE Journal 8(2), 149–156.
  11. IPCC (2005). IPCC Special Report on Carbon Dioxide Capture and Storage (Metz, B. et al., eds.). Cambridge Univ. Press, 442 pp.
  12. Bachu, S. (2015). Review of CO₂ storage efficiency in deep saline aquifers. International Journal of Greenhouse Gas Control 40, 188–202.
  13. Peaceman, D.W. (1978). Interpretation of Well-Block Pressures in Numerical Reservoir Simulation. SPE Journal 18(3), 183–194.
  14. Theis, C.V. (1935). The relation between the lowering of the piezometric surface and the rate and duration of discharge of a well using ground-water storage. Trans. AGU 16, 519–524.
  15. US EPA (2010, 2013). Federal Requirements Under the UIC Program for CO₂ Geologic Sequestration Wells (40 CFR Parts 124, 144, 145, 146, 147 — Class VI); Geologic Sequestration of Carbon Dioxide: Underground Injection Control (UIC) Program Class VI Well Area of Review Evaluation and Corrective Action Guidance, EPA 816-R-13-005.
  16. Alberta Energy Regulator (2017). Directive 065: Resources Applications for Oil and Gas Reservoirs. AER, Calgary, AB.
  17. Aziz, K., Settari, A. (1979). Petroleum Reservoir Simulation. Applied Science Publishers. — IMPES, transmissibility, numerics.
  18. Ennis-King, J., Paterson, L. (2005). Role of convective mixing in the long-term storage of carbon dioxide in deep saline formations. SPE Journal 10(3), 349–356. — Convective (Rayleigh) dissolution flux.

Engine identity for the audit trail: SubsurfaceAI VE / 3D Layered Engine · pure-JS implementation · MIT-style license · lineage MRST-co2lab (academic open source). Git: github.com/tchisti/Nisku_CO2_Simulator.