Heat Recovery Justification and a Slipping Project

Simple payback (undiscounted): cumulative savings are 480, 970, 1,469, 1,979 and 2,498 (thousand $) at the end of years 1–5. Payback = 4 + (2,400 − 1,979)/520 = 4.8 years, about half the evaluation period - a reasonable first screen, but it ignores the time value of money and the escalation.

NPV. Savings grow at g = 2% and are discounted at r = 8%, so use a growing annuity:

PV = A/(r − g) × [1 − ((1+g)/(1+r))^n] = 480/0.06 × [1 − (1.02/1.08)^10] = 8,000 × [1 − 0.5645] = 8,000 × 0.4355 = $3,484k

NPV = 3,484 − 2,400 = +$1,084k. On a 10-year, 8% basis the project is clearly worth doing.

Downside test. A 15% saving reduction gives PV = 0.85 × 3,484 = $2,961k and NPV = +$561k. Still strongly positive. The break-even is a fall in savings of 2,400/3,484 = 31%, which is more than twice the 15% downside the consultant considers plausible. Note also that the break-even is where NPV = 0, not where the project fails economically - a 31% shortfall leaves the company indifferent on NPV but still better off than not investing.

Recommendation. Proceed. The decision is robust to the identified downside, and the sensitivity that matters more than energy price is capital cost and the commissioning date: a 10% capex overrun ($240k) wipes out about a fifth of the NPV, and every month of delay costs roughly $40k of deferred savings. That points the management attention at schedule and cost control rather than at the energy price assumption.

Derived indices (month 8):

• CV = EV − AC = 1,400 − 1,720 = −$320k
• SV = EV − PV = 1,400 − 1,600 = −$200k
• CPI = EV/AC = 1,400/1,720 = 0.814
• SPI = EV/PV = 1,400/1,600 = 0.875
• Physical progress = 1,400/2,400 = 58% complete.

Forecast final cost - present a range, not a single number:

• Efficiency-based (remaining work at the same CPI): EAC = BAC/CPI = 2,400/0.814 = $2,948k, VAC = −$548k.
• One-off-cause basis (the overrun is a discrete event that will not recur): EAC = AC + (BAC − EV) = 1,720 + 1,000 = $2,720k, VAC = −$320k.
• Combined cost-and-schedule basis: EAC = AC + (BAC − EV)/(CPI × SPI) = 1,720 + 1,000/0.712 = $3,124k.

The spread of roughly $400k is the honest answer at this point: the data cannot yet distinguish a systemic productivity problem from a one-off. The sponsor should be told that the project is 58% complete, has consumed 72% of the budget, and that the expected outcome is $2.7–3.1 million.

Presentation points.

1. The cost variance is already 19% of earned value; a variance of that size at 58% complete rarely closes on its own. Assume at least the $2.72M figure for funding purposes and hold the difference as management reserve.
2. Do not let the disputed $180k change order be used to explain the CPI before it is adjudicated. If the AC includes work performed under an unapproved change, exclude it and recompute - otherwise the performance picture is distorted in both directions.
3. SPI 0.875 on a six-month build is about three weeks of schedule erosion, and it is compounding with the two-month heat-exchanger delay. Since the exchanger is almost certainly on the critical path, commissioning at month 7 becomes month 9, deferring roughly $80k of savings and exposing the project to liquidated damages if the contract provides for them.

Establish entitlement first, quantum second.

1. Gather the record: tender drawings and specifications, the issued-for-construction set, the RFI log, the vendor skid general-arrangement drawings and their revision dates, and the minutes of any design meetings. The question is whether the steel is additional to the tendered scope or merely the contractor's under-estimate of a scope they were required to allow for.
2. Test the cause. If the steel became necessary because the fan skid vendor changed the support points after tender, that is an owner-side design development and the change is compensable. If it arises because the contractor's tender omitted bracing shown on the tender drawings, it is contractor risk and the claim should be rejected, with the reasoning in writing.
3. Test quantum. Even where entitled, ask for a quantity build-up (tonnage, rate per tonne, connection allowance, painting, design fees) and benchmark the rate against the original steelwork bill. Lump-sum change orders should not be accepted without a build-up.
4. Check for betterment and duplication. If the additional steel also serves a future expansion or a heavier skid than specified, deduct the betterment.

Schedule-related cost to set against it. The change order should be assessed together with the two-month heat-exchanger delay, not in isolation. Deferred commissioning of two months defers savings of about 2/12 × $480k = $80k in first-year money, and slightly more once the escalation profile is applied. If the additional steel is itself on the critical path, the two delays may not be additive - but if the steel work is needed before the exchanger can be set, it could extend the outage further. Set the $80k (plus any liquidated damages that the contract allows) against the $180k: the true incremental cost of the owner's position may be roughly $100k, not $180k, and that changes the negotiating posture.

Recommendation. Approve only the entitled portion, in writing, with a clear reservation of rights on delay; instruct the contractor to proceed with the steel under a instructed-variation letter so the critical path is not held hostage to the commercial negotiation. Where a change is small relative to the cost of delay, paying it and pressing the schedule is the better commercial outcome - but only after the cause is documented.