Reduce Manufacturing Demand Charges with Interval Data Analysis.
Your commercial electricity bill is not just about total consumption—it is heavily driven by a single 15-minute peak usage interval each month. The software analyses your utility meter data to identify the specific equipment causing your peak demand, and calculates the financial value of rescheduling those loads.
Two numbers say whether you have a problem
Two numbers from your last bill are all I need to determine if you have a viable opportunity for cost reduction. You get your answer back within a day. If there is no potential, I will tell you directly.
The line nobody reads
A commercial manufacturing power bill consists of two parts: total energy used (kWh) and peak demand (kW). The demand charge often accounts for up to a third of total electricity costs.
Because utilities bill based on your highest 15-minute average load across the entire month, a single simultaneous machine start-up can dictate your rate for all 30 days.
The software helps you lower this cost by identifying how to sequence your load—without reducing your operational output. You do not have to use less electricity; you just have to stop using so much of it at once.
How much room you actually have
Average load divided by peak load. A plant running flat around the clock sits near 100% and has almost nothing to gain. I will tell you that rather than sell you an analysis. A plant at 16% is paying all month for a peak that lasted minutes, and that is where the money is.
What that looks like in numbers
350 kW machine shop, PG&E B-10, load factor 29%
- Annual electricity bill
- $319,000
- Of which demand charge
- ~28%
- Peak reduction from rescheduling
- 60–90 kW
- Annual saving
- $17,500 – $25,500
- My fee
- a share of it, agreed before I start
Nothing about production changes. Same output, same shifts, same trucks. The starts move by thirty to sixty minutes.
Modelled on PG&E's published 2026 tariff (Advice 7846-E). An illustration of the method. Not a client result.
How it works
- 01
Initial screening
Provide your total monthly kWh and billed peak kW. I calculate your load factor and reply within 24 hours to advise if a full analysis is financially viable for your facility.
- 02
Data analysis
If the initial numbers show promise, you provide a year of your utility interval data. The software processes this data to pinpoint the exact 15-minute intervals setting your peak price.
- 03
Rescheduling model
I generate a highly specific, data-backed schedule (e.g., staggering machine start times) modeled to prevent demand spikes. The software is constrained to only produce schedules that mathematically lower your modeled peak.
- 04
Monthly verification
I compare your new billing data against a baseline model of your unoptimized usage to verify the financial impact month over month.
What it costs
The screening and the full interval-data analysis are free, and so is the number that comes out of them. If I find nothing worth doing, you pay nothing and that is the end of it. If we do find something, the fee is a share of what I found rather than a rate per hour, and we agree it after you have seen your own number. You keep the larger part of it, every year, for as long as the schedule holds.
Who this works for
Worth a look
- Commercial bakeries and food production with batch ovens, spiral mixers, or a hard morning ramp.
- Fruit, vegetable and nut processing facilities with seasonal, sharply peaked loads.
- Wineries and breweries managing crush and bottling on top of steady refrigeration.
- CNC machining and metal fabrication.
- Plating, anodizing, powder coating, and heat treating facilities where rectifiers hammer the demand meter.
- Commercial printing operations.
Probably not
- 24/7 flat-rate production lines with no schedulable processes.
- Consistent cold-storage facilities holding one temperature all year.
- Flat-rate tariffs with no demand component at all.
- Any facility with a load factor above roughly 60%.
We would rather turn you down in one email than take a fee for finding nothing. The two numbers tell us which list you're on before anyone spends real time.
Who you're dealing with
Relay Studio operates as a focused team. The analysis software is built internally, not licensed or resold.
It also means saying no to plants where the numbers don't work. That's the whole reason the two-number check exists.
Every client-facing figure is checked against the utility's published tariff book. The software refuses to produce one from a rate schedule nobody has verified. I work with PG&E customers today and read the rates straight off your own bill everywhere else.
The studio is registered in Austria. The mechanism is the same everywhere. California is where it costs the most.
Reasonable questions
- Do I have to change what I produce?
- No. The whole method is about when loads run, not how much they draw. Same output, same shifts. If a process can't move, three ovens that all have to be on at four because the truck leaves at six, the analysis shows that and I tell you.
- What if you don't find anything?
- You've lost thirty seconds and I've lost an afternoon. The first analysis is free and nothing is attached to it.
- What happens to my meter data?
- It stays in local storage or an encrypted vault. Never in a public repository, never published or shown to anyone without written permission. Ask us to delete it and it's deleted.
- How do I know the saving is real?
- I measure against a model of what the bill would have been, cross-check that against the physical mechanism, and bill on whichever is lower. If the statistical error on a month is too big to defend, we don't claim it.
- Why should I trust a modelled number?
- You shouldn't, which is why every figure on this page is labelled as modelled. The number that matters is the one computed from your own interval data against your own tariff, and that's what the free analysis produces.
Thirty seconds to find out
Two numbers off your last bill. An answer within a day. No call needed.