← Back to Insights
Your energy bill hides a second bill.
"Demand charges alone account for up to 40% of an industrial facility's electricity bill. They are charged on peak consumption minutes, not average usage. A microgrid does not just produce energy. It restructures when and how you consume it. That distinction is where the P&L case lives. "
40%
Of an industrial electricity bill can be demand charges, priced on your peak minute, not your average usage.
(ABB industrial microgrid benchmark, 2024)
(ABB industrial microgrid benchmark, 2024)
3-7 yrs
Typical payback period for commercial microgrid and battery storage systems in 2026.
(GSL Energy, 2026)
(GSL Energy, 2026)
THE DECISION STAKES
The utility grid charges you for what you consume in 15 minutes.
""The business case is really cost management. This is how you make money with the microgrid: you want to manage your cost."" — Dr. Michael Stadler, Co-Founder and CTO, Xendee, Energy Central, May 2025
A microgrid is not an energy project. It is a demand management instrument that also provides resilience and carbon reduction. The P&L case must model all three value streams.
Peak shaving alone can justify the investment in high-demand-charge environments. Resilience value, carbon cost avoidance, and grid services revenue shorten the payback further. The analysis that ignores these underestimates the return by 30 to 50%.
THE DECISION TOOL
Four moves. One decision you can defend.
01
AUDIT
Analyse 12 months of interval energy data: identify demand charge peaks, duration and frequency of grid events, and carbon intensity by hour. The audit reveals where the hidden bill lives and what size of system eliminates it.
Without interval data, you are sizing a microgrid on averages. Averages hide the 15-minute peaks that generate 40% of your bill.
02
SIZE
Model three scenarios against your load curve: peak shaving only, full islanding for resilience, and a hybrid that optimizes for all value streams. Match technology selection (solar, battery, CHP) to your specific site constraints and grid tariff structure.
Oversizing reduces flexibility and increases payback. Undersizing misses demand charge savings. The optimal size is the one your interval data reveals, not the vendor's standard configuration.
03
FINANCE
Stack every available value stream in the financial model: demand charge reduction, energy arbitrage, grid services revenue, tax credits (ITC/MACRS), and carbon cost avoidance. Evaluate Power Purchase Agreement versus ownership. The deal structure determines whether payback is 3 or 7 years.
US federal investment tax credit at 30% and MACRS depreciation together reduce net project cost by up to 40%. In the EU, similar incentive structures apply. The model that ignores these is not conservative. It is incomplete.
04
OPERATE
Define the operating protocol before go-live: who manages dispatch decisions, how demand response contracts interact with islanding needs, and what triggers grid export vs. storage. An unmanaged microgrid captures 60% of available value. A managed one captures 90%.
The operating protocol is the difference between a system that pays back in 5 years and one that takes 9. It costs nothing to define before deployment and everything to retrofit after.
UCSD Energy Services
University of California San Diego. 42 MW microgrid capacity, 62 MW peak. One of North America's most studied industrial-scale microgrids.
92%
Of campus electricity generated on-site, eliminating exposure to grid demand charges and peak pricing. The microgrid includes solar, fuel cells, CHP, and thermal storage, managed by a central energy management system.
(UCSD Energy Services, public data)
UCSD runs one of the most complex and best-documented microgrids in North America. The 42 MW system covers a campus consuming more electricity than a small city. By generating 92% of its electricity on-site and managing demand in real time, the campus eliminates the demand charge exposure that accounts for the largest share of conventional utility bills. The system islands from the main grid during outages, maintains 99.9%+ uptime, and generates tens of millions of dollars in annual savings compared to equivalent grid-supplied electricity. The business case was built entirely on cost management, not sustainability.
UCSD did not build its microgrid for resilience. It built it because 92% self-generation was cheaper than paying demand charges to the grid.
Download the full case
PDF · 5 slides · Free access · Downloaded 0 times
Key questions
At what ratio of peak-to-average demand does the microgrid business case become self-financing without grid services revenue?
How do you model the resilience value of islanding capacity in financial terms for a board that only approves capex with a defined payback?
When does a Power Purchase Agreement outperform ownership on a 10-year NPV, and which site characteristics determine the crossover?
Pre-decision checklist
AUDIT — completed
SIZE — completed
FINANCE — completed
OPERATE — completed
By Fabrice Macarty
This case resonates?
Audit 12 months of interval data before sizing anything. Model peak shaving, resilience, and grid services as three separate lines. Stack tax credits into the financial model before committing. Define the operating protocol before the first switch is thrown.
Start the conversation