Most energy cases are lost in the first two minutes, before anyone mentions strategy. The interviewer is not waiting for a profitability tree; they are waiting to hear whether you treat nameplate megawatts as a revenue forecast. That single habit — multiplying MW by a price as if the plant ran every hour of the year — is what separates a passable structure from a case that never recovers.
After this piece you should be able to convert capacity to energy with a real capacity factor, put LCOE next to contracted and merchant prices, and decide a build or offtake question the way a score sheet actually rewards.
What the score sheet rewards first
Energy practices (strategy arms of Big Four deals shops, MBB energy teams, economic consultancies that live in rate cases) give these prompts because the units punish sloppy candidates. Upstream oil, a gas peaker, a renewable PPA, and an LNG offtake are different businesses. They share one grading sequence.
| Order | What they listen for | Typical fail |
|---|---|---|
| 1 | Commodity and market design (power, gas, oil, LNG; regulated vs merchant) | Jumping to "cost cut" without naming the revenue calculus |
| 2 | Capacity vs energy (MW vs MWh) | Revenue = MW x hours x price with no capacity factor |
| 3 | Revenue stack (PPA, merchant capture, capacity payment, regulated ROE) | Mixing a contracted $/MWh with a pool average |
| 4 | Contractual fixed costs (take-or-pay, ship-or-pay, reservation charges) | Putting take-or-pay in variable COGS |
| 5 | Only then: dispatch, fuel, carbon, grid, curtailment | Opening with a factory-style fixed/variable tree |
I stop listening when a candidate says "volume times price" and the volume they wrote down is still in MW.
The four metrics that show up in real data
Interviewers expect you to sound like someone who has seen a plant model, not a Wikipedia page. Anchor on primary ranges, then treat case exhibits as law.
Capacity factor. Annual energy is approximately nameplate MW x 8,760 hours x capacity factor. For U.S. utility-scale fleets in 2024, EIA State Energy Data System capacity factors (Table F53) report about 23.2% for solar PV, 34.3% for wind, 60.5% for natural-gas combined cycle, and 13.9% for gas turbines (peakery territory). If the case gives you a different factor, use the case. If it does not, ask for it before you forecast MWh.
LCOE. Levelized cost of energy folds capital recovery, fixed and variable O&M, fuel (if any), and financing into a $/MWh that is comparable across technologies — with the caveat that LCOE is not the same as value of energy delivered in a given hour. Lazard's LCOE+ Version 17.0 (June 2024) puts unsubsidized onshore wind at about $27–73/MWh, utility-scale solar at $29–92/MWh, gas combined cycle at $45–108/MWh, and gas peaking at $110–228/MWh. Globally, IRENA's 2024 cost report puts weighted-average new onshore wind near $34/MWh and utility solar near $43/MWh (2024 dollars). Use these as sanity checks; the case's own LCOE wins.
Merchant vs contracted revenue. A PPA (or other offtake) locks a $/MWh (or a capacity payment). Merchant capture is what you actually clear in the wholesale market after shape, curtailment, and negative correlation with other wind or solar. They are not interchangeable. Large developers still prefer contracted books where they can: NextEra Energy's 2024 Form 10-K states that as of year-end 2024 roughly 94% of NEER's net generating capacity sat under long-term contracts, with shorter hedges used when spot exposure remains (NextEra Energy 2024 Form 10-K).
Regulated returns. In a rate-base utility case, equity earnings are closer to rate base x equity ratio x allowed ROE than to "MWh times pool price." The same NextEra 10-K describes FPL's authorized regulatory ROE midpoint at 10.80% (band 9.80%–11.80%) under the settlement triggered by Treasury rates. If the prompt is a regulated build, open with allowed return and capital employed, not a merchant spark spread.
| Business type | Revenue math that belongs on the page |
|---|---|
| Merchant generator | MWh x captured price (+ capacity market if any) |
| Contracted IPP | MWh x PPA price (watch must-take, shape, curtailment clauses) |
| Regulated utility | Rate base x equity % x allowed ROE (+ riders) |
| Midstream / offtake | Throughput fees and take-or-pay reservations as fixed costs |
Units that fail candidates before strategy starts
MW vs MWh. MW is an instantaneous rate. MWh is energy. Almost all energy revenue lines are MWh (or a separate capacity payment in $/MW-year). Writing "150 MW plant at $40/MWh" without a capacity factor implies 100% utilization — about 1.31 TWh and $52m of fantasy revenue for a plant that may deliver a quarter of that.
Heat rate and fuel. Thermal variable cost is roughly heat rate (MMBtu/MWh) x fuel ($/MMBtu) plus variable O&M and carbon. Do not invent a conversion from gas to power unless the case gives a heat rate.
Take-or-pay / ship-or-pay. You pay for reserved capacity whether or not you use it. That is a fixed cost with a contract attached, not optional volume. Spare contracted volume only has value if you can divert or trade it at a positive netback.
Curtailment and interconnection. MWh you cannot deliver are not MWh you can sell. Ask for curtailment or congestion when the project is behind a constrained node.
Worked mini-case: Cedar Ridge Wind
Prompt (invented company and figures). Cedar Ridge Energy, an independent power producer, holds development rights for a 180 MW onshore wind project in the Midwest. The board wants a commercialize-or-kill recommendation this week. Exhibits give:
| Input | Case value |
|---|---|
| Nameplate | 180 MW |
| Capacity factor | 35% |
| All-in project LCOE | $46/MWh |
| Corporate PPA offer | $51/MWh fixed, 12 years, must-take when available |
| Expected merchant capture | $34/MWh (high-wind hours clear soft) |
| Pipeline take-or-pay (affiliate ask) | 50,000 MMBtu/day reservation at $0.45/MMBtu; expected use 22,000 MMBtu/day for a peaker that is not FID'd |
Two decisions: PPA or merchant for the wind? Sign the pipeline reservation?
Step 1 — Energy, not capacity.
Annual MWh = 180 x 8,760 x 0.35 = 551,880 MWh.
The trap answer multiplies 180 x 8,760 x $51 and quotes about $80m of PPA revenue. That assumes a 100% capacity factor. With 35%, you are in a different decade of the P&L.
Step 2 — Contracted vs merchant vs LCOE.
| Path | Revenue | Implied LCOE cost | Contribution vs LCOE |
|---|---|---|---|
| PPA | 551,880 x $51 = $28.15m | 551,880 x $46 = $25.39m | +$2.76m |
| Merchant | 551,880 x $34 = $18.76m | $25.39m | -$6.62m |
The PPA clears the project's own LCOE with a thin but positive spread. Merchant does not. Unless the case gives a credible path for capture to rise above LCOE (storage, different node, capacity payments), take the PPA. Lazard's onshore wind band ($27–73/MWh) and EIA's ~34% U.S. wind capacity factor make these case inputs look interview-plausible; they are still case law, not a market forecast.
Step 3 — Take-or-pay as deadweight until diversion exists.
Annual reservation = 50,000 x 365 x $0.45 = $8.21m.
Unused capacity = (50,000 - 22,000) / 50,000 = 56%, about $4.60m/year of fixed cost with no FID'd offtake. Do not dress that up as a "strategic option." Reject the pipeline reservation until the peaker has committed volume or a trading book with documented diversion netbacks.
Recommendation. Sign the wind PPA; decline the pipeline take-or-pay. Risks to flag: offtaker counterparty strength (ask for a parent guarantee), curtailment at the interconnection, and merchant temptation if the PPA has an early termination clause. Next exhibits to request: hourly capture vs PPA shape, curtailment history at the node, and any capacity-market revenue the project can still stack.
Ned's rule. In an energy case, convert MW to MWh with a capacity factor before you say the word "revenue." If contracted price is below LCOE, or merchant capture is below LCOE with no credible uplift, you are arguing about how to lose money more slowly — say that out loud.
Regulated builds are a different opening tree
When the client is a regulated utility, swapping wholesale price for allowed ROE is not optional color; it is the case. A simplified check: if the exhibit shows $900m of rate-base addition, 50% equity ratio, and a 10.8% allowed ROE, annual equity earnings from that addition are about 900 x 0.50 x 0.108 = $48.6m, subject to the usual regulatory lag and prudence fights. Compare that to a merchant alternative on the same capital only after you have named both calculators. The NextEra 10-K language on FPL's ROE band is useful context for what "allowed return" looks like in a real U.S. settlement; the interviewer's exhibit still governs.
What a generic profitability tree misses
- Capacity is not energy; utilization lives in the capacity factor (and in curtailment).
- Contracts create fixed costs that look like "volume" on a slide.
- Shape and capture, not the annual average pool price, set merchant revenue.
- Fuel and carbon dominate thermal variable cost and barely appear for wind or solar until you add storage.
- Regulated ROE and merchant $/MWh answer different questions; mixing them is how candidates invent phantom synergies.
If you treat Cedar Ridge like a factory at full utilization, you will roughly triple the energy, sign a reservation for molecules you do not need, and still feel proud of the framework.
Practice this on real unit economics
Run one timed math pass where you force yourself to write MW, capacity factor, MWh, price path, and contribution on separate lines before any "so what." CoachNed's quick math and case math drills are built for that habit; everything is open for seven days with no card, then $120 for a recruiting season or $49 a month.
For a full voice case after the units click, use /interview. For industry variety after energy, the case library at /consulting-cases includes operations and other sector prompts.
Frequently asked questions
What is the difference between MW and MWh in a case interview?
MW is power (a rate). MWh is energy (rate times time). Revenue and LCOE almost always sit on MWh, while capacity payments may sit on MW-year. Always ask for a capacity factor before turning nameplate into annual energy.
What capacity factor should I assume if the case does not give one?
Ask. If forced to estimate for U.S. utility-scale context, EIA's recent annual figures are roughly mid-20s percent for solar PV, mid-30s for wind, around 60% for gas combined cycle, and low-to-mid teens for gas peakers — then label the estimate and invite the interviewer to correct it.
Is LCOE the same as the price I should use for revenue?
No. LCOE is a cost benchmark. Revenue is PPA price, merchant capture, capacity payments, or regulated return. A project can have a low LCOE and still lose money on merchant capture if it sells into soft hours.
How do I handle take-or-pay in an energy case?
Treat unpaid-for unused volume as a fixed cost. Only keep the contract if expected use plus diversion or resale covers the reservation at a positive netback. "Strategic optionality" without a netback is usually a soft pass on the score sheet.
Do McKinsey, BCG, and Bain use real energy numbers in cases?
They use interview-style exhibits. Public LCOE and capacity-factor ranges from Lazard, EIA, and IRENA are useful for sanity checks and for sounding numerate, but the case numbers on the page override any external benchmark. CoachNed is independent and not affiliated with McKinsey, BCG, Bain, or any firm named for practice.
Sources
- EIA State Energy Data System, Table F53 capacity factors (2024) — U.S. utility-scale solar PV (~23.2%), wind (~34.3%), natural-gas combined cycle (~60.5%), and gas-turbine (~13.9%) capacity factors. Checked 2026-09-24.
- Lazard LCOE+ Version 17.0 (June 2024) — Unsubsidized LCOE ranges for onshore wind, utility solar, gas combined cycle, and gas peaking. Checked 2026-09-24.
- IRENA Renewable Power Generation Costs in 2024 — Global weighted-average LCOE context for new onshore wind and utility-scale solar. Checked 2026-09-24.
- NextEra Energy Form 10-K for the year ended December 31, 2024 — NEER contracted capacity share (~94%) and FPL authorized regulatory ROE (10.80% midpoint). Checked 2026-09-24.
- IEA Renewables 2024 — Broader renewable deployment and system-integration context for interview discussions of grids and growth. Checked 2026-09-24.