Commercial Demand Charges and the 300 kWac Interconnection Threshold
A food-distribution warehouse has a 480 V service, 400 kW peak demand, 1.2 GWh/year, and a demand ratchet of 80% of the highest winter peak for 12 months. The utility offers net billing: exports are credited at avoided energy cost (~$0.045/kWh), not retail, and demand charges are based on the highest 15-minute interval each month. The owner wants a 500 kWdc / 400 kWac PV system to "eliminate the bill." The service transformer is 500 kVA and the existing main breaker is 600 A. Minimum daytime load is 180 kW; peak load occurs at 6 p.m. in summer. The roof is available, but the utility requires an interconnection study for systems above 300 kWac and may require a relay/SCADA. The owner asks whether to size at 300 kWac to avoid study delays or go larger.
1. Estimate how much of the 400 kWac PV output will be self-consumed and how much exported; what does that mean for bill savings?
Read the worked answer
Start with annual energy: 400 kWac × about 1,500 kWh/kWac = 600 MWh/year, against a 1,200 MWh load. Annual energy alone suggests 50% of the load, but the bill impact depends on timing. The minimum daytime load is 180 kW, so on low-load days the array can exceed site load and export. On normal weekdays, mid-day load may be 250–350 kW, so much of the output is self-consumed. A reasonable planning range is 70–85% self-consumption and 90–180 MWh of exports, but this must be confirmed with 15-minute interval data and hourly production modeling.
Value the energy separately. If 500 MWh is self-consumed at an all-in energy rate of about $0.10/kWh, that is $50,000/year. If 100 MWh is exported at $0.045/kWh, that is only $4,500/year. The export is worth less than half of retail, so adding AC capacity mainly to export is a weak investment.
Demand charges are the bigger issue. The summer peak occurs at 6 p.m., when PV output is low or zero, so the array will not reduce that peak. The winter ratchet is set by the highest winter peak, which may occur on a cloudy morning. PV may reduce some monthly peaks, but it will not eliminate demand charges. The answer to "eliminate the bill" is no; the proposal should show energy savings, export credit, and only the demand reduction that coincides with the load peak.
2. Interconnection threshold: what are the trade-offs of 300 kWac vs 400 kWac, including study, equipment, and project schedule?
Read the worked answer
At 400 kWac, the project crosses the utility’s study threshold. Expect a formal interconnection study, possible feeder or transformer upgrades, relay/SCADA requirements, and a longer schedule-often several months beyond a smaller project. Study and equipment costs can be tens of thousands of dollars, and the utility may assign network upgrade costs. The 500 kVA transformer and 600 A main may be adequate, but that is a utility determination, not a sales assumption.
At 300 kWac, the project may avoid the formal study and relay/SCADA scope, but do not assume it avoids review entirely; confirm the utility’s exact threshold and screening rules. The production loss from 400 to 300 kWac is not proportional because the 500 kWdc array can be oversized relative to the inverter. With a 500 kWdc/300 kWac ratio, clipping may lose roughly 5–8% of annual production, but the lower AC rating reduces interconnection risk.
Compare net present value, not just gross production. The extra 100 kWac may add about 120–150 MWh/year; at a blended value of $0.07/kWh, that is $8,400–$10,500/year. If the study and upgrades cost $40,000 and delay the project by nine months, the incremental project may still work financially, but the risk-adjusted return may be worse than a 300 kWac build. Present both options with schedule, upgrade risk, and production sensitivity.
3. The owner wants a proposal with "no demand charges." How do you correct that and what data do you need?
Read the worked answer
Explain that demand charges are set by the highest 15-minute average demand in each billing month, sometimes with a ratchet. PV reduces demand charges only when it is producing during the interval that sets the peak. A summer peak at 6 p.m. is after the PV window, so the array cannot reduce that peak. A winter peak on a cloudy morning may also escape reduction. A ratchet can carry a high winter peak for 12 months, so one bad interval affects the whole year.
Request 12 months of 15-minute interval data, the complete tariff including demand ratchet and power factor rules, and any minimum-demand charges. Model PV production at the same 15-minute resolution and calculate the net load for each interval. Then produce a before/after bill for each month, not an annual average. Include a sensitivity case with clouds and a case with a new load.
If demand reduction is a real goal, discuss battery storage or load shifting, but size it against the actual peak interval. Do not write "no demand charges" in the proposal. Write "estimated demand-charge reduction based on modeled coincidence" and show the assumed interval data. That language is accurate and still gives the owner a clear financial picture.