Foundry SEC Spike and Audit Project Ranking
An induction-melting foundry reports 11,000 MWh for the year at 18,000 t of saleable castings. The approved baseline year made 22,000 t with 12,000 MWh, but 3,000 MWh of that year was fixed holding-furnace, compressed-air and lighting load; the rest varied with tonnage. The energy manager has three proposals: (A) furnace waste-heat recovery to preheat charge, 1.2 GWh/yr saved, capex Rs 4.5 crore; (B) variable-frequency drives on cooling-water pumps, 0.6 GWh/yr, Rs 1.1 crore; (C) scheduling and ladle-preheat discipline, 0.4 GWh/yr, Rs 15 lakh. A capacitor bank is also proposed, but it reduces kVA demand only. Production is expected to stay near 18,000 t next year, with a 10% chance of a 2,000 t export order. The plant wants a defensible audit recommendation for the board.
1. Separate fixed and variable load, calculate the production-normalized baseline and excess energy, and state whether the reported SEC rise is real.
Read the worked answer
Baseline variable load = 12,000 − 3,000 = 9,000 MWh. At 22,000 t, variable SEC = 9,000/22,000 = 0.4091 MWh/t. Expected energy at 18,000 t = 3,000 + 0.4091 × 18,000 = 10,364 MWh. Actual energy is 11,000 MWh, so excess = 636 MWh, about 6.1% above expected. The reported SEC = 11,000/18,000 = 0.611 MWh/t, but the correct comparison is normalized baseline SEC = 10,364/18,000 = 0.576 MWh/t. The apparent jump from 0.545 to 0.611 is mostly a fixed-load effect; only 636 MWh is genuinely unexplained. Before crediting projects, investigate that 636 MWh through submetering, furnace holding time, compressed-air leaks and scheduling. This is the audit finding that should go to the board.
2. Rank A, B and C on annual kWh saved, capex per annual kWh, simple payback at Rs 7/kWh, and delivery risk. What do you recommend?
Read the worked answer
Capex per annual kWh: A = Rs 4.5 crore/1.2 GWh = Rs 37.5/kWh; B = Rs 1.1 crore/0.6 GWh = Rs 18.3/kWh; C = Rs 15 lakh/0.4 GWh = Rs 3.75/kWh. At Rs 7/kWh, annual savings are A Rs 84 lakh, B Rs 42 lakh, C Rs 28 lakh; simple paybacks are A 5.4 years, B 2.6 years, C 0.54 years. C is first because it is low-cost, fast and tests operating discipline. B is second: lower capex per kWh and shorter payback than A. A is largest but highest capex and depends on charge-preheat integration; it should follow a measurement phase. The capacitor bank gives no kWh saving, so it belongs in a tariff/PF decision, not in the kWh ranking. If the 2,000 t export order occurs, variable load rises; the fixed-load projects retain value, but re-run the ranking with 20,000 t.
3. What M&V protocol and audit evidence would make the savings claim defensible?
Read the worked answer
Use IPMVP Option B for the pump VFDs and waste-heat recovery: isolate the retrofit, meter before/after, and adjust for production. For scheduling, use Option A with stipulated operational savings plus logs. Establish a 12-month baseline with 15-minute meters on furnace, pumps, compressor and lighting; record production, shift pattern and ambient temperature. For the normalized baseline, preserve the fixed/variable regression and state uncertainty. For A, meter flue-gas and charge temperatures, heat recovered and fan power; for B, pump head/flow and VFD frequency; for C, ladle-preheat times and furnace idle hours. Report savings net of the 636 MWh unexplained gap so the board does not double-count. Include tariff and demand savings separately.