The completion and stimulation program: capstone

Three months into the rotation, Ethan presents the completion and stimulation program for the Tamarack 4-22H horizontal well to Tom, Lena and the operations manager. It is the program for the next horizontal well in the Bakken, in the neighboring spacing unit, using everything the team learned from the 3-22H and from the earlier lessons of this course. This lesson walks through his program section by section, the way a real program is written and reviewed.

1. Objectives

The program opens with objectives the whole team can check. For the 4-22H: deliver an estimated ultimate recovery of at least one million barrels from a 9,800 ft lateral; keep completion cost under about 5.5 million dollars; use at least a third recycled water; complete the job with no lost-time injuries, no spills to ground and no seismic event above the amber threshold; and collect the diagnostic data needed to improve the next well.

2. Well and reservoir data

The second section collects what the design rests on. The lateral will be landed in the middle Bakken at about 10,450 ft true vertical depth, drilled along the minimum horizontal stress direction. The stress model of lesson 7 gives a closure gradient of about 0.77 psi/ft, about 8,040 psi at the landing depth, to be confirmed by a DFIT in the toe stage. Pore pressure is about 0.51 psi/ft. The lateral has a cemented 4-1/2 in P-110 liner, and fracturing will be done down a 4-1/2 in P-110 tieback frac string with a burst rating of about 12,400 psi, which sets an 80 percent treating pressure limit of about 9,900 psi and protects the 7 in intermediate casing. A cement evaluation log is required before perforating. Offset wells within a mile are identified for monitoring and notification.

3. Completion design

Ethan proposes cemented plug-and-perf with dissolvable plugs, which avoids a coiled tubing drill-out run in a long lateral. He considers two designs. The base case repeats the 3-22H: 40 stages of five clusters at about 245 ft stage length. The alternative uses 50 stages of five clusters, about 195 ft long, with the same in-basin 40/70 and 100 mesh sand. Lesson 9 showed that conductivity is not the constraint here, so the alternative spends money on more fractures, not stronger proppant. Perforating uses five equal-entry holes per cluster, lesson 10's limited-entry result.

4. Treatment design

Each stage is pumped at 80 bbl/min of slickwater with a high-viscosity friction reducer, with a small acid spearhead on the first stages if breakdown is difficult. The schedule has a pad of about a fifth of the stage volume, sand ramped from 0.25 to 1.0 lbm/gal, and a flush that stops short of the perforations. The base case uses about 6,500 bbl of fluid and 250,000 lbm of proppant per stage. From the stage records of the 3-22H, Ethan adds more pad and lowers the final concentration on the toe stages, which screened out on the first well. The expected surface treating pressure is about 9,100 psi, under the limit but with only a modest margin, so he recommends close monitoring and a lower rate if pressures run high. The pumps need about 17,800 hydraulic horsepower.

5. Comparing the designs

# Capstone: comparing two completion designs on cost, recovery and water logistics
SI = False  
if SI:   # 03-22 Cardium horizontal; thousand Canadian dollars, m3
    designs = {  # name: (stages, clusters per stage, proppant t, fluid m3, EUR m3 of oil)
        "base: 20 stages x 4 clusters, sand":          (20, 4, 600.0, 5200.0, 36300.0),
        "tighter: 26 stages x 4 clusters, sand + RCS": (26, 4, 830.0, 6800.0, 44500.0),
    }
    cost_fixed, cost_stage, cost_prop, cost_fluid = 180.0, 31.9, 0.16, 0.012     # per stage incl. spread time (lesson 1), per t, per m3
    netback, cu, vu, truck = 0.33, "thousand CAD", "m3", 30.0                       # thousand CAD per m3 of oil; m3 per truck load
else:    # 3-22H Bakken; thousand US dollars, bbl
    designs = {
        "base: 40 stages x 5 clusters, sand":          (40, 5, 10.0e6, 260000.0, 942000.0),
        "tighter: 50 stages x 5 clusters, sand":       (50, 5, 13.0e6, 330000.0, 1040000.0),
    }
    cost_fixed, cost_stage, cost_prop, cost_fluid = 420.0, 66.8, 0.00006, 0.0025   # per stage incl. spread time (lesson 1), per lbm, per bbl
    netback, cu, vu, truck = 0.040, "thousand USD", "bbl", 130.0
base_cost = base_eur = None
for name, (n, c, prop, fluid, eur) in designs.items():
    cost = cost_fixed + n * cost_stage + prop * cost_prop + fluid * cost_fluid
    value = eur * netback
    print(f"{name}")
    print(f"   cost {cost:7,.0f} {cu}; EUR {eur:9,.0f} {vu}; undiscounted net value {value - cost:8,.0f} {cu}; water {fluid:,.0f} {vu} = {fluid/truck:,.0f} truck loads if trucked")
    if base_cost is None: base_cost, base_eur = cost, eur
    else:
        d_cost, d_eur = cost - base_cost, eur - base_eur
        print(f"   extra cost {d_cost:,.0f} {cu} for {d_eur:,.0f} {vu} more oil: {d_cost/d_eur*1000:,.0f} dollars per {vu} added, against a netback of {netback*1000:,.0f}")
print("the recovery uplift is a forecast with wide uncertainty; a pilot on two or three wells tests it before the whole program changes")

base: 40 stages x 5 clusters, sand
   cost   4,342 thousand USD; EUR   942,000 bbl; undiscounted net value   33,338 thousand USD; water 260,000 bbl = 2,000 truck loads if trucked
tighter: 50 stages x 5 clusters, sand
   cost   5,365 thousand USD; EUR 1,040,000 bbl; undiscounted net value   36,235 thousand USD; water 330,000 bbl = 2,538 truck loads if trucked
   extra cost 1,023 thousand USD for 98,000 bbl more oil: 10 dollars per bbl added, against a netback of 40
the recovery uplift is a forecast with wide uncertainty; a pilot on two or three wells tests it before the whole program changes

The base design costs about 4.3 million dollars and is expected to recover about 942,000 bbl, matching the 3-22H. The alternative costs about 5.4 million dollars and is forecast to recover about 1.04 million bbl, 98,000 bbl more for about 1.0 million dollars, or about 10 dollars per extra barrel, against a netback of about 40 dollars. It also uses about 330,000 bbl of water, about 2,500 truck loads if trucked, which strengthens the case for piping water to the pad. The economics favor the alternative, but the uplift is a forecast from offset well data, not a certainty, and closer fractures may interfere more with the neighboring wells. Ethan recommends the alternative, with fiber optics in the well to measure cluster efficiency, so that the next wells can be designed on evidence.

Lena, the reservoir engineer, challenges one assumption: whether the extra recovery is new oil or oil that the neighboring well would have produced anyway. Ethan agrees that the type-curve uplift may be partly acceleration rather than additional reserves, and adds a sensitivity case in which only half the uplift is real. The alternative still pays, by a smaller margin. This is the kind of question every program should face before approval, because completion designs that look good on paper can take oil from offset wells instead of adding it.

6. Operations plan

The operations section covers the sequence and the people. The well will be completed about two months after rig release, zipper-fractured with the sister well on the pad. Steps: rig up and pressure test the wellhead, frac tree and tieback string; run the cement evaluation log; perforate the toe with coiled tubing and run the DFIT; fracture the toe stage; then plug-and-perf the remaining stages around the clock, alternating with the sister well; flow back through a sand separator and test separator; pull the tieback string and install tubing with gas lift as described in lesson 2. Water comes by pipeline from a permitted depot and from the field's produced water recycling, stored in tanks on secondary containment. Sand comes from in-basin mines and is stored in enclosed silos with dust control.

7. Safety, environment and stakeholders

The safety plan identifies the main hazards: high pressure in the treating iron, explosives in perforating, hydrogen sulfide during flowback, respirable silica, heavy traffic and lifting. Controls include the exclusion zone around the high-pressure lines, the perforating procedures of lesson 3, gas detection, silica controls and a traffic management plan. The environmental plan covers spill prevention, emissions control with flowback gas routed to sales or a flare, noise screening near a farmstead, and the company's induced seismicity traffic-light protocol. Regulatory steps include the required notices to the Industrial Commission, the offset operator notifications, water permits and the fracturing fluid disclosure.

8. Evaluation plan

The final section says how the team will know whether the design worked. Fiber optics in the well will show cluster efficiency during the job; tracers by stage group will show contribution along the lateral; offset pressure gauges will record any frac hits; and the first six months of production will be compared with the 3-22H and other offsets using rate-transient analysis normalized for lateral length and proppant. A short report after six months will recommend the design for the following wells.

The review

The operations manager asks the questions every completion engineer should expect. What happens if the DFIT shows higher stress than the model? The rate comes down and the limited-entry design is rechecked. What if a screenout occurs? The stage is flowed back if possible or cleaned out with coiled tubing, and the following stages get more pad. What if a magnitude 2.0 event occurs? The amber response follows, with the plan's reduced-rate stages. What if the well underperforms? The diagnostics will show whether the problem is the rock, the clusters or the proppant, and the next design changes accordingly.

Tom's last comment closes the course. 'You've written a program a field crew can follow and a manager can approve. Remember the order: understand the rock, protect the well, place the fluid and sand where they belong, keep people and the land safe, and measure what you got. Every well you complete teaches you how to complete the next one.'

A checklist for any completion program

Objectives that can be checked. Well and reservoir data: depth, stresses, pore pressure, rock and fluid properties, casing and cement records. Completion type and its justification. Perforating design. Stimulation design: fluids, proppant, schedule, rates, pressures and their limits. Tubing and upper completion. Operations sequence, equipment and logistics. Safety, environment, neighbors and regulatory steps. Cost estimate and economic case with sensitivities. Evaluation and data collection plan. Contingencies.

Approval and handover

Once approved, the program becomes the reference for everyone involved. The service companies prepare their own detailed procedures and job designs that must match it; any change in the field, such as a lower rate or a different proppant, is recorded with its reason, and changes that affect safety or pressure limits need approval from the engineer in charge. During the job, the operator's completion engineer in the data van checks each stage against the program and decides whether to continue, adjust or stop.

After the job, the completion report records what was actually done: every stage's rate, pressure, volumes, proppant placed, events and observations, the perforation and plug depths, the cost and the safety record. That report is what Ethan read on his first morning, and it is what the next engineer will read before designing the next well. Writing it carefully is the last step of the completion, and one of the most valuable.

Ethan's program is approved with two changes: a lower maximum treating pressure on the first five stages until the DFIT confirms the stress, and an extra offset pressure gauge in the nearest producing well. Both make the program safer and the evaluation better, and both came from the review, which is why programs are reviewed.