Primary depletion combines gas and oil expansion, solution gas, aquifer and gas cap drives to produce hydrocarbons from a saturated field without any enhancements (such as waterflooding) [6]. Primary depletion of a gas cap reservoir derives main production energy from gas cap expansion, and solution gas drive, with slow gas cap-aided reservoir pressure reductions resulting in higher production for longer [6].

Water flooding is an often inefficient, but inexpensive secondary recovery technique in saturated reservoirs [10], where water displaces oil from the pore space and drives it towards the producing wells [11]. Oil recovery volumes vary depending on factors such as oil viscosity, petrophysics, and the natural reservoir drive mechanisms, as these all affect the water’s ability to displace and drive the oil [10]. Timing of the flooding is key as the earlier it is begun, the minimised the primary depletion becomes, thus the limited the gas saturation increase is (as higher gas saturation decreases oil recovery) [11]. Well, location is critical to maximised water flooding, with three common arrays used—peripheral/edge drive, line drive and 5-spot (Figure 6) [11]. The tighter the well spacing, the better the recovery and total sweep efficiency (E_T, the effectiveness of the injected water to “sweep” oil into the producers), which is affected by both local-to-well (E_L) and field-wide, areal factors (E_A), predominantly reservoir rock properties [6]:

E T = E A ∗ E L

The more favourable reservoirs for waterflooding are those that: are shallower (cheaper costs and often lower primary recovery); have a lower Bo (resulting in lower gas saturations and lower primary recovery); have higher permeability (to maximise water flow and utilise wider well spacing) [11].

Like water flooding, gas injection is the dispersion of injected gas into the reservoir, either from the beginning, or when pressure begins to drop. Injection is typically done directly into the reservoir, to both increase pressure and decrease oil viscosity, but can also be done directly into the gas cap, to maintain gas cap drive mechanisms [6]. The issue with this is the added cost of an extra injector well being drilled directly into the gas cap, which adds both significant extra cost, and can lead to coning [6].

If economically viable, gas produced from a saturated field can be re-injected to aid oil production (re-cycling). However, this requires more surface facilities, such as separators and compressors, increasing costs, as well as the potential production impact from the initial gas production [6].

During drilling of any field, there is limited data available regarding the reservoir properties, consequently it is crucial to account for variations in the accuracy of the data to develop acceptable assumptions regarding the field’s economic output. By identifying key reservoir properties, and analysing their impacts on primary production rates and totals using Schlumberger Eclipse, economic analysis can be carried out to ascertain the worst case (P90), most likely/base case (P50), and best case (P10) scenarios for the economics in terms of the field’s primary recovery NPV (Table 2).

Within the saturated reservoir in this study, the location/size of the gas cap, via the location of the gas-oil contact (GOC), residual oil saturation (S_or), permeability and porosity were outlined as four critical properties that must be known for accurate economic analysis to be completed. This study varied a sole property at a time and ran oil production simulations to assess their impacts, after which economic analysis was run on each of the simulations (Table 3).

The results show that varying reservoir porosity has the largest impact in production totals, with the P90 and P10 scenarios (porosity of 16% and 24%, respectively) resulted in recovery of 31753.79 × 10³ STB and 47582.60 × 10³ STB, respectively, a variation of approximately 15.8 million STB. In comparison, permeability fluctuations have relatively negligible effect on recovery totals, with the P90 and P10 scenarios (kv/kh of 5/50md and 15/150md, respectively) producing 39337.63 × 10³ STB and 39694.52 × 10³ STB, respectively, a variation of approximately 0.35 million STB (Table 4).

Both GOC location and S_or show one scenario veering significantly from the base case (P50) and one approximately tracking along it. The GOC P90 (gas cap of 125 ft) shows a significantly lower recovery than both the P50 (100 ft gas cap) and P10 (75 ft gas cap), whereas the residual oil saturation P10 (15%) recovery total is significantly higher than both the P50 (25%) and P90 (35%). These variations, from both the GOC and S_or, and the porosity, means that further investigation into the reservoir properties would be highly beneficial, providing the gathering of the data is economically viable (Figure 8).

The sensitivity analysis ran on the primary depletion simulations, using P50, P90 and P10 variables, produced the NPV’s for each scenario (Table 5). These values clearly show how different properties impact recoverable totals, and how they also impact NPV. Each property impacts the NPV to a different degree, with Figure 9 outlining to what extent, within this project. The figure compares the NPV of each scenario against the base case, most likely NPV outcome (P50), from this, analysis can be undertaken to assess the impact of each property on the field’s potential value. This analysis took several assumptions regarding costs, outlined in Appendix 1.

As the tornado diagram indicates, the NPV from each scenario reflects the production total variations outlined within the simulations. The gas cap thickness sees a $78.28 mm decrease in NPV for the P90, but only a $9.23 mm increase for the P10 upside, inferring that a gas cap greater than 50 ft thick can have a vast impact on total recovery, whereas a gas cap less than 50 ft has a smaller impact on NPV. On the other hand, this is reversed for residual oil saturation, where a P90, 35% S_or is only $36.23 mm below the baseline, compared to the P10, 15% S_or, which is $99.79 mm greater than the baseline NPV. Comparatively to both of these, porosity has the most extreme impacts on NPV, but with both P90 and P10 having approximately the same effect ($149.05 mm and $149.17 mm, respectively). Opposite to porosity is the impact permeability has on NPV, with both the P90 and P10 showing minimal variations in NPV, whilst both being approximately equal ($19.25 mm and $14.26 mm, respectively).

This analysis suggests that the need for further information to reduce these large variations in assumptions, and thus large variation in NPV, is generally required, though this is further investigated in the following VOI analysis.

The VOI is used to carry out cost/benefit analysis on EMVs for potential further investment-scenarios, in this case, for another appraisal well (costing $10 mm) to gain full knowledge on a reservoir property. To do this, each reservoir property must be analysed for their EMV, and if a profit can be created through narrowing the assumptions of that property (i.e. if the cost of a new appraisal well is cheaper or costlier than the change in EMV caused by the P90/P10 variations in the property) (see Appendix 2).

As Table 6 shows, gas cap thickness, residual oil saturation and porosity all have monetary benefits to another appraisal well (VOI’s of $11.88 mm, $24.01 mm and $64.56 mm, respectively). However, the minimal variations in permeability NPVs and EMVs in each sensitivity scenario is reflected in a VOI of −$1.62 mm, inferring that the knowledge, and thus monetary benefits gained from another appraisal well are outweighed by the cost of the well itself. Yet, as the other three properties would greatly benefit from another well, this project recommends the drilling of one in order to maximise monetary value and knowledge of the reservoir.

This project assessed the viability of water flood enhanced oil recovery through numerical simulations and economic modelling. Primary depletion, taken from the sensitivity analysis testing, was used as a benchmark, and was carried out alongside simulations with 1 - 4 injector wells. This project used a daily water injection rate of 24,000 bbl per injector. The fourth injector showed both limited production increases, and was limited by the number of producers available in the reservoir (also four), so a fifth injector was not investigated. An educated trial-and-error method was employed to identify optimum location of the injector wellsto maximise sweep efficiency, and thus to increase the final cumulative recovery value. When the maximum production total was identified, that well array was deemed the optimum arrangement for the relative well amount.

As Table 7 and Figure 10 show, there is a clear increase in production due to the enhanced recovery from water injection, whilst water flooding also maintains higher production rates for longer.

Gas injection into a saturated reservoir has one of two purposes, depending on the planning—injection into the gas cap to improve gas cap drive, and injection into the oil-producing zone to maintain reservoir pressure. This comparison is one of the many decisions involved in economic analysis. Like the waterflooding simulations, gas injection was conducted in an variety of injector numbers (1 - 4), all in a various array of wells, in attempts to maximise sweep efficiency, with the highest-producing array chosen for economic analysis. This simulation used a daily gas injection rate of 24,000 SCF per injector. The limit of four injectors was, again, caused by field limitations (four producing wells) and limited increases in production between three injectors and four. This gas injection investigation plotted the range of gas injectors against both primary depletion and waterflooding models, as well as one and two gas cap gas-injectors, with economic comparisons to be analysed in detail later in this project.

As Table 8 shows, gas cap injection into the field has a minimal improvement on recovery compared to primary depletion, suggesting that this is not a viable option due to the gas cap properties. However, gas injection directly into the oil phase does offer improved recovery techniques (Figure 11). These, like waterflooding, appear to plateau at 3 - 4 injectors. However, initial interpretations of the data highlights that two gas injector wells do not reach the total produced from a single water flood system.

Initial results showed that any plateau rate set at ±1000 STB/day from the minimum value would not allow enough production build-up time (seen within the 9000 STB/day plateau rate aggregation).In the spreadsheets used, the first variable tested was production rates, beginning at 6000 STB/day, therefore the minimum plateau rate was set at 9000 STB/day.

In solution gas drive, plateau rates of 9000 STB/day gave high NPV values of $499.33 mm, regardless of the well timing and the capped cumulative production rate became closer to the potential cumulative rate. Very high plateau rates, above 50,000 STB/day, resulted in very high facilities costs adding to the OPEX and resulting in low NPVs and PI.

However, the best well timing from this report was every 3 months (quarterly), this suggest that early investment in new wells means less expenditure on facilities in the long run. Delaying well development leads to increased costs due to inflation, reducing OPEX, but demanding an initially higher CAPEX than other options. Overall, the recovery factor of solution gas drive is 39.09% this is higher than the waterflooding at 28%. Solution gas drive is a better development option than waterflooding, as it gave a higher NPV of $499.44 mm compared to $291.22 mm. This was achieved at the optimal plateau rate of 9000 STB/day and well interval of 3 months for solution gas drive, while waterflooding was at 6 months and 9000 STB/day (Figure 12).

From the optimization of the saturated oil field, the best development option was based on the PI and not NPV as money is a limitation. PI of 1 is logically the lowest acceptable measure for the index, any lesser value renders the project useless and it’s abandoned. Therefore, they would develop solution gas drive as it has the highest PI of 1.64 (Table 10).

Long well intervals are not desirable as yearly intervals gave the lowest NPV and productivity index regardless of the plateau rates and facilities cost. However, well interval of 12 months and plateau rate of 24,000 STB/day gave the worst NPV because of the cost of installing new wells increases due to inflation as time goes by, thereby adding to the facilities cost and having the same production rate causes a decline in the NPV. Therefore, cases of yearly well intervals were ignored. Since the discovery well flowed at 6000 STB/day and start the of plateau usually occurs 2 to 10 years after build up time, using plateau rate below 9000 STB/day is not considered as optimizing the well. As plateaus rate from all four producing wells would exceed 6000 STB/day. Hence, the minimum plateau rate was set at 9000 STB/day for each case. Early investment in new wells means less expenditure on facilities in the long run, give the best NPV value during development planning. As can be seen above, 3 months well interval gave the highest NPV and plateau rates of 9000 to 24,000 STB/day, making this the ideal optimal well timing and rate for the field.

Solution Gas Drive is a better development option than waterflooding, as it gave a higher NPV of $499.44 mm compared to $291.22 mm (Table 11). Although the facilities cost for solution gas drive is 20% more expensive per STB/day of oil than waterflooding due to an extra cost of drilling 4 injector wells, an additional profit of $208.22 mm was made in the solution gas drive case.

During the optimization processes, the best plateau rate for solution gas drive case is 9000 STB/day and still lies in the well time of every 3 months as it gave the highest NPV. So, every 3 months, investments are made on 1 producer well and 1 injector well three times to give the total number of 8 drilled wells. Plateau rates below 15,000 STB/day for waterflooding gave a significantly lower NPV value and rate, as can be observed from Table 1; all the plateau rates of 9000 bbl/day gave an NPV less than $290 mm (Figure 13).

The sole purpose of waterflooding is to maintain reservoir pressure and increase sweep efficiency, however, this newly discovered saturated field’s initial pressure is 6000 psi, which is well above the bubble point pressure of 3500 psi. However, it will be a wise option to postpone the use of waterflooding and rely solely on energy from the expansion of the rock and fluids until the pressure drops below the bubble point and then drill an injector well to increase and maintain the reservoir pressure. This will increase the CAPEX and the economics of the project.