Option C Savings Claim Confounded by a New Tenant and a Server Room
An energy savings performance contract covers a chiller replacement, VFDs on two 50 hp supply fans, and an EMS upgrade in a 220,000 ft² office building. The contract uses IPMVP Option C with a 12-month baseline and a 3-year term. Base year: 4,850,000 kWh from utility bills, 5,120 HDD, 1,850 CDD. Year 1: 4,210,000 kWh, 5,890 HDD, 1,610 CDD. The ESCO invoices 640,000 kWh of savings. During year 1 the owner extended operating hours for a new tenant (about 18% more occupied hours) and installed a 40 kW server room now served from the house meter. The contract says only that 'savings shall be determined per IPMVP' and contains no adjustment clause. The owner's finance office has withheld the year-1 payment.
1. Is the 640,000 kWh claim defensible as invoiced? Identify every adjustment that is required and quantify what you can.
Read the worked answer
No - the claim is not defensible as invoiced. 4,850,000 - 4,210,000 = 640,000 kWh, which is a raw bill subtraction with no normalization at all.
Adjustment 1 - weather (works against the ESCO in a heating-dominated year). HDD rose 5,890/5,120 = +15%; CDD fell 1,610/1,850 = -13%. A simple degree-day ratio is invalid here because a large share of 4.85 GWh is weather-independent base load (plug, lighting, elevators, DHW). The correct method is a regression of monthly kWh against HDD and CDD (3- or 4-parameter change-point model) fit to the baseline. With only 12 monthly points the model is at the IPMVP/FEMP minimum - it should have been 24–36 months.
Adjustment 2 - non-routine: the server room. 40 kW x 8,760 h = 350,400 kWh/yr of pure IT load at the meter, before counting the cooling energy to reject it (roughly another 30–40%). That is more than half the entire claimed savings, and it is added load - it must be added to the adjusted baseline, not counted as a saving.
Adjustment 3 - non-routine: 18% more occupied hours. Fans, pumps, lighting and plug loads all scale with occupied hours. With the VFD fans alone, 18% more run hours partially offsets the VFD savings.
Conclusion: after a proper weather-normalized baseline plus the server-room and occupancy adjustments, the true savings are plausibly near zero or negative. The withholding is likely justified; the ESCO's number is an accounting artifact of a cold winter and a contract that never defined adjustments.
2. The contract is silent on adjustments. What measurement approach should the project have used, and what would you put in the M&V plan to make year 2 and 3 enforceable?
Read the worked answer
Option C was the wrong choice here. Whole-building metering is only appropriate where occupancy, tenancy and loads are stable. This building changed tenant mix and added a large 24/7 load in year 1.
Recommended structure:
- Option B (retrofit isolation) for the chiller plant: dedicated kW meter on the chiller(s) plus measured ton-hours (flow and delta-T, or BMS points validated against a portable meter). Savings = baseline kW/ton x measured ton-hours minus measured post-retrofit kWh.
- Option B for the two fans: VFD kW transducers, or BMS kW with a quarterly spot-check. Fan savings are then insulated from tenant behavior.
- Keep Option C as a cross-check only, never as the payment basis.
Minimum M&V plan content:
- Baseline regression with stated model form, R², and CV(RMSE) - target ≤ 5% for whole-building, and 24–36 months of bills.
- A written non-routine adjustment procedure: what events trigger adjustment (change in conditioned area > 5–10%, occupancy change > 10%, new process loads, hours change), who calculates it, the deadline, and dispute resolution.
- The server room sub-metered separately and placed outside the measurement boundary entirely.
- Meter calibration certificates, data-quality/backup clauses, and a defined missing-data protocol.
- Stated uncertainty and a stated consequence if uncertainty is exceeded (e.g., savings trued up to the lower bound).
Practical fix for the current dispute: convert year 1 to an Option B calculation using the chiller and fan meters, and settle the difference; then amend the contract before year 2.
3. The owner now wants to 'take back' the savings and self-verify while paying the ESCO a fixed fee. What are the risks, and what would you require before agreeing?
Read the worked answer
Risks of take-back plus self-verification:
- Incentive inversion. Once the ESCO is on a fixed fee, nobody is paid to commission, re-optimize, or fix a drifting economizer for three years. Persistence is the part of an ESPC that usually erodes fastest; a fixed fee removes the mechanism that protects it.
- Loss of independence. Self-measurement by the party that also wants the savings number to look good is the classic conflict. Adjustment disputes (like the one already underway) become unenforceable.
- Financing exposure. If the project is financed against the savings stream, an unverifiable savings number threatens the debt service and the lender's acceptance.
- Baseline drift. With no third party, no one will document the new tenant, the server room, or the next change.
What I would require before agreeing:
- A written IPMVP-compliant M&V plan signed by both parties, with Option B metering on the chiller plant and fans.
- Meters installed and calibrated by a third party; calibration certificates and raw data files turned over to the owner.
- Annual independent verification by an accredited verifier (AEE CEM/CEM-adjacent or CMVP), paid from an escrow, not by either party's discretion.
- A defined baseline-adjustment procedure and dispute-resolution ladder, with a 30-day cure period.
- A retained performance bond or holdback tied to the verified savings in years 2 and 3, so the ESCO still has skin in the game even on a fixed fee.
- A commissioning re-check at months 12 and 24 - trending economizer, schedule, and static pressure reset - because those are where savings quietly disappear.