EOR economics and capstone

A year after Lena's question, Ethan presents his recommendation to the Prairie Crest management committee in Calgary. He has screened the Cardium A pool, worked through the laboratory results with Mei Lin, rebuilt the waterflood design, diagnosed the thief layer with the field team at Drayton Valley, and run the history-matched model for every process that passed the screen. This capstone lesson follows the last step: turning the technical work into an economic comparison and a staged plan that a company can act on. It draws on every earlier lesson, and it shows that the best EOR process is not always the one that recovers the most oil.

What makes EOR economics different

EOR projects share several features that set their economics apart from drilling new wells. First, they are incremental: the value of the project is the difference between the forecast with EOR and the forecast without it, so the waterflood base case must be forecast just as carefully as the EOR case, and both come from the same model. Second, they are front-loaded: injection facilities, recycle plants, pipelines and well conversions are paid for before the oil response arrives, typically one to three years later. Third, they carry a large and continuing injectant cost, whether CO2, polymer or surfactant, which behaves like an operating cost per barrel and makes the project sensitive to oil price. Fourth, they are long: a CO2 or chemical flood runs for 15 to 30 years, so the discount rate strongly affects the answer.

The usual measures are those of PET-450: net present value at the company's discount rate, internal rate of return, payout time and capital efficiency, expressed as NPV per unit of capital invested. For EOR two further measures are common. The injectant cost per incremental barrel, the net utilization of Lesson 7 multiplied by the injectant price, is the quickest way to compare processes and fields. The breakeven oil price is the price at which the NPV is zero; projects with a breakeven well below the company's planning price are robust.

Screening the pool: a summary

Three options reach the economic comparison. The first is a CO2 WAG flood, with average pressure raised about ten percent above the slim-tube MMP of Lesson 6, a tapered WAG ratio near 1:1 and 0.6 HCPV of CO2, forecast to recover about 12 percent of the original oil in place. The second is a polymer conformance flood aimed at the thief layer, using a deep-placed gel at the worst injectors and then polymer at about 10 mPa·s (10 cp), forecast to recover about 3 percent. The third is a surfactant-polymer flood, forecast to recover about 10 percent but with the retention and salinity risks of Lesson 10.

# Capstone economics: three EOR options for the pool, with NPV, payback and a pilot decision.
import math
SI = True      # True: Canadian case (metric reporting), False: US case (oilfield reporting)
BBL = 6.2898
ooip_bbl = 5.68e6 * BBL                         # original oil in place, bbl
price = 90.0 if SI else 70.0                    # dollars per bbl
roy = 0.15 if SI else 0.25                      # royalty and production taxes
co2_cost = 40.0                                 # dollars per tonne delivered
credit = 0.0 if SI else 30.0                    # storage credit passed to the operator, dollars per tonne
rate = 0.10

opts = {   # name: (fraction of OOIP, delay yr, tau yr, capex M, opex dollars/bbl, injectant dollars/bbl)
    'CO2 WAG':              (0.120, 1, 4.0, 80.0, 14.0, 0.32 * (co2_cost - credit)),
    'polymer conformance':  (0.030, 1, 3.0, 15.0, 10.0, 4.0),
    'surfactant-polymer':   (0.100, 2, 4.0, 70.0, 14.0, 22.0),
}
def profile(frac, delay, tau, years=20):
    sh = [max(t - delay, 0) * math.exp(-max(t - delay, 0) / tau) for t in range(1, years + 1)]
    return [ooip_bbl * frac * s / sum(sh) for s in sh]

res = {}
for name, (frac, d, tau, capex, opex, inj) in opts.items():
    oil = profile(frac, d, tau)
    cf = [-capex * 0.6e6, -capex * 0.4e6] + [0.0] * 18
    for t, q in enumerate(oil):
        cf[t] += q * (price * (1 - roy) - opex - inj)
    npv = sum(c / (1 + rate) ** (t + 0.5) for t, c in enumerate(cf))
    cum, pay = 0.0, None
    for t, c in enumerate(cf):
        cum += c
        if pay is None and cum > 0: pay = t + 1
    res[name] = npv
    vol = f"{sum(oil) / BBL / 1e6:.2f} million m³" if SI else f"{sum(oil) / 1e6:.2f} MMbbl"
    print(f"{name:20s} oil {vol:16s} NPV10 {npv / 1e6:6.1f} M, NPV/capex {npv / (capex * 1e6):4.2f}, payout year {pay}")

# Pilot first? CO2: pilot 8 M, success probability 0.6; failure after a full project loses 70 % of capex
p, pilot = 0.6, 8.0e6
go_now = p * res['CO2 WAG'] + (1 - p) * (-0.7 * opts['CO2 WAG'][3] * 1e6)
with_pilot = -pilot + p * res['CO2 WAG'] / 1.1 ** 2
print(f"CO2 full project now, expected value {go_now / 1e6:5.1f} M; pilot first {with_pilot / 1e6:5.1f} M")
print("Currency: " + ("Canadian dollars" if SI else "US dollars"))

CO2 WAG              oil 0.68 million m³  NPV10   36.4 M, NPV/capex 0.46, payout year 6
polymer conformance  oil 0.17 million m³  NPV10   25.0 M, NPV/capex 1.66, payout year 4
surfactant-polymer   oil 0.57 million m³  NPV10    4.1 M, NPV/capex 0.06, payout year 8
CO2 full project now, expected value  -0.5 M; pilot first  10.1 M
Currency: Canadian dollars

The results frame the decision. The CO2 flood has the largest NPV, about 36 million Canadian dollars, but it also needs the most capital and pays out only in its sixth year. The polymer conformance project is much smaller but has by far the best capital efficiency, with an NPV of about 1.7 times its capital and payout in year four. The surfactant-polymer flood barely breaks even once its chemical costs are counted, and that is before its technical risk is considered. The expected-value calculation adds a further point: with a 60 percent chance of technical success, committing to the full CO2 project now has an expected value near zero, while a pilot first, at 8 million dollars, raises the expected value to about 10 million by avoiding a large loss if the pilot fails.

Risk and the value of a pilot

Every EOR forecast carries large uncertainty: the sweep in the thief layer, the response time, the injectant utilization, and the oil price over 20 years. The PET-420 forecast for the waterflood pilot already had a P10 roughly twice its P90, and an EOR forecast built on top of it is wider still. A field pilot is a way of buying information. It costs money and time, but it can reveal a fatal flaw, such as early CO2 breakthrough through the thief layer or unacceptable surfactant retention, before the full investment is made. Pilots must be designed to answer specific questions: an inverted five-spot or a pair of patterns with observation wells, tracers and logging, and a clear definition in advance of what result would justify expansion.

Phasing is the other risk tool. A project can be developed pattern by pattern, starting where the reservoir is best understood and the surface facilities closest, so that early patterns' performance guides later ones and early cash flow helps fund expansion. Phasing reduces peak capital exposure and makes good use of recycle capacity, at the cost of some delay in the full benefit.

Ethan's recommendation

Ethan's recommendation has three stages. First, start the polymer conformance programme now: gel treatments at the three injectors connected to the thief layer, followed by polymer injection in the affected patterns. It is cheap, pays out quickly, and improves the sweep that any later process will rely on. Second, raise the average pool pressure above the MMP by running the waterflood at a voidage replacement ratio near 1.1, and negotiate for captured CO2 from the Alberta Carbon Trunk Line system or another nearby source. Third, run a two-pattern CO2 WAG pilot with an observation well and tracers in the best-characterized part of the pool, with expansion conditional on CO2 breakthrough later than 0.3 HCPV and net utilization below 2.5 tonnes per cubic metre. Surfactant-polymer flooding stays in the laboratory until retention can be cut, and low-salinity water is tested in cores alongside it.

Lena's comment after the meeting is the last word of the course. “You did not recommend the process with the biggest number. You recommended the one we can afford to be wrong about, and you set up the test that tells us whether the big one is real. That is what EOR engineering is.” The committee approves the conformance programme and funds the CO2 supply study.

Sensitivities and the breakeven price

Before the meeting, Lena asks Ethan to test how robust each option is. He reruns the economics with oil price 20 percent lower, injectant costs 20 percent higher, incremental recovery 30 percent lower and a one-year delay in response, one at a time. The incremental recovery has the largest effect for every option, followed by oil price. The CO2 flood is the most sensitive to price because its injectant cost is fixed per tonne while revenue falls; the polymer conformance project is the least sensitive because its capital is small and its response is quick. The surfactant-polymer flood loses most of its value under any single adverse change.

He also computes the breakeven oil price of each option by searching for the price at which the NPV is zero. A short loop over prices in steps of one dollar is enough, and the result is easy to communicate: a project whose breakeven price is well below the company's long-term planning price can survive a downturn, while one whose breakeven is close to it is a bet on price. Ethan's chart of breakeven prices becomes the single slide the committee discusses longest.

Two further points come up in discussion. The first is timing: the value of an EOR project depends on starting it while the waterflood infrastructure, wells and pipelines are still in good condition and before the pool's production falls to its economic limit. Waiting has a cost, because the base case declines and wells are abandoned, and a project that is marginal today may be impossible in ten years. The second is portfolio fit: a large CO2 project would take most of the company's capital for two years, while the conformance programme leaves room for drilling elsewhere. Both points favour the staged plan.