Page 60

Page 61

Graphical representation of the probable range of results in the different scenarios. A) Summary of intermediate 
and final volumes for scenario ‘All’, and the theoretical and effective volumes for scenario ‘All’ and scenarios ‘A–E’. The 
error bars signify the P90–P10 range from the plotted mean value. For scenarios ‘All’ and ‘B’, the fracture contribu-
tion is also shown on the same graph for comparison. The two deterministic CO2 requirement volumes (1.2 million 

tons and 0.2 million tons), using the CO2 density end-members, are plotted for comparison. Note the logarithmic 

scale. GRV gross rock volume, NRV net rock volume, PRV pore rock volume. B) Pie chart illustrating the contribution 
of the three stratigraphic zones and the fracture system to the overall porosity of the system. c Variance diagram 
illustrating the critical parameters with largest impact on the total volume. Scenario ‘All’ does not have any seg-
ments and thus highlights parameters that are most uncertain and should be addressed through a scenario-based 
calculation. From Senger et al. (2015)

Forecasting phase relations in the Longyearbyen CO2 lab reservoir 

           - 

Miri 

et 

al. 

2014

Understanding of fluid mixture properties relevant to the Longyeabyen CO2 Lab pilot project will impact the injec-
tion performance. Phase equilibria and density of the binary, ternary and quaternary systems containing CO2, CH4, 
H2O and NaCl have been investigated using a statistical associating fluid theory (SAFT) based equation of state 
(EOS) at ambient temperature and pressure and salt concentrations up to 5.5 mol/kgH2O, all relevant to LYBCO2. 
Binary interaction parameters of the subsystems (CO2-CH4, CH4-H2O, and CH4-NaCl) were tuned against available 
experimental data, using previously adjusted parameters for pure components and CO2-H2O subsystems. Solubility 
of CH4 and CO2 and subsequent mixture densities were predicted at 298 K and pressure up to 100 bar. It is found that 
by increasing the hydrocarbon in the injection stream (even in small amount) and also salt concentration, solubility of 
the CO2 in the aqueous phase and consequently the density of the mixture will reduce. Moreover, hydrocarbon impu-
rities like CH4 would result in a favorable density difference and faster plume migration, however the probability of 
three phase state (two liquid and one vapor phase) near the bubble line is very high too. The results of this work are 
applicable to estimation of storage capacity as well as simulation of plume migration and fate in all projects facing 
CO2, CH4, H2O, and NaCl bearing fluid system. 

Sketch of the Longyearbyen CO2 Lab (Svalbard), illustrating the framework for CO2 injection into reservoir sand-

stones at 670–970 m depth. The forecasted CO2 plume will interact with groundwater and possibly methane gas. 

The explored stratigraphic succession is shown in the column to the right, with the top-seal given by the Janusfjellet 
Subgroup and the reservoir by the Kapp Toscana Group. The upper 120–150 m of the near-surface overburden is in 
the permafrost, shown with blue colour in the sketch. From Miri et al. (2014).