Maximum Demand Charge: How It Is Calculated, Why It Ratchets, and What Actually Reduces It
Maximum demand charge is billed on the single highest average kVA your meter recorded in any one integration block during the billing month — commonly a 15-minute or 30-minute block — not on your average load and not on the units you consumed. One block, once a month, sets a charge you pay for the entire month, and under a ratchet clause it can set a floor for the months that follow.
That is why a plant drawing a steady load all month can still be billed as though it ran flat out. A compressor restarting while a chiller is loaded and a mill is being started is enough. The energy charge is about how much you used; the demand charge is about how badly you spiked. They behave differently, and only one of them responds to the things most plants try first.
What the meter is actually recording
Your HT meter does not record an instantaneous peak. It integrates power over a fixed block — the integration period — and stores the average for each block. At the end of the month, the highest of those block averages is your recorded maximum demand.
Maximum demand is the highest block average recorded in the billing period, expressed in kVA for most industrial tariff categories.
This matters more than it sounds. A one-second inrush when a 250 kW motor starts direct-on-line barely moves a 15-minute average. A furnace charging for twelve minutes of a fifteen-minute block moves it a long way. The integration period separates a transient you can ignore from a load you must manage, so find out which period your tariff uses first.
Contract demand, recorded demand and billing demand are three different numbers
These three get used interchangeably in meetings, and the confusion costs real money.
Contract demand is the capacity you have contracted with the utility to draw. It drives your sanctioned load, your connection arrangement and often the minimum you pay regardless of use.
Recorded demand is what the meter actually measured in the highest block of the month.
Billing demand is the number the charge is applied to. It is derived from the other two by a rule written in the tariff order in force — typically the higher of the recorded demand and some stated fraction of the contract demand, with a penal treatment applied above contract demand.
| Term | Set by | Changes when | Where you see it |
|---|---|---|---|
| Contract demand | Your agreement with the utility | You apply and the application is sanctioned | Agreement copy, and printed on the bill |
| Recorded maximum demand | Your own plant, block by block | Every month, with your worst block | Demand line on the monthly bill |
| Billing demand | The tariff order's formula | Only when the order changes or your demand changes | The line the demand charge is computed on |
| Excess demand | Drawing above contract demand | Any month you overshoot | Usually a separate penal line |
Do not assume the formula. Read the clause in the order that applies to your tariff category, because the definition of billing demand is the single clause with the greatest influence on your bill.
Why the charge is in kVA and not kW
Your plant does work in kW. The utility has to build and maintain the network that carries kVA. Cables, transformers and switchgear are sized by current, and current follows apparent power, so the utility charges for apparent power.
kVA is kW divided by power factor. A plant drawing 800 kW at 0.85 power factor presents 941 kVA to the network; the same 800 kW at 0.98 presents 816 kVA. Same useful work, 125 kVA of difference in what you are billed on.
This is why power factor correction and demand reduction are often the same project. It is also why a plant with poor power factor is paying twice: once in the demand charge computed on inflated kVA, and again in any power factor penalty the tariff applies. The thresholds and the amounts are set by order and change, so read the order in force rather than repeating a figure someone quoted you years ago.
What the ratchet clause does
This is the part nobody reads until it bites.
A ratchet or minimum-billing clause sets your billing demand to a stated fraction of the highest demand recorded in a defined past window, even in months when you drew far less. The window may be the preceding financial year or a rolling number of months, and the fraction is stated in the tariff order.
The consequence is simple and expensive. One bad block can carry a cost forward through every quiet month that follows. The plant that ran its whole shutdown at a fraction of normal load, then restarted everything at once on the final morning, pays for that morning for a long time.
Our recommendation is that restart after any shutdown is a controlled sequence with a written order of switching, not a race to get the plant running. Stagger the compressors. Bring chillers up before the process heat load. A PLC-based load management scheme exists for exactly this.
How to read your own bill in five minutes
Take the latest bill and find these lines. If your utility uses different wording, the concepts still map.
| Bill line | What it means | What to do with it |
|---|---|---|
| Contract demand / sanctioned demand | Your contracted capacity in kVA | Compare with twelve months of recorded demand; chronic under-use is money wasted |
| Maximum demand recorded | Your worst block this month | Plot all twelve months; look for the outlier, not the average |
| Billing demand | What the demand charge was computed on | If it exceeds your recorded demand, a ratchet or minimum clause is operating |
| Demand charge | Billing demand multiplied by the rate in the order in force | The rate is set by tariff order and changes; never budget on last year's figure |
| Power factor / kVArh data | Average power factor or reactive energy | A penalty or incentive line, and a direct input into your kVA |
| Excess demand charge | Drawing above contract demand | Recurring excess means your contract demand or your load control is wrong |
| Units consumed (kWh) | Energy, not demand | This is the number solar reduces; demand is a separate fight |
Twelve months of this on one sheet tells you more than any audit presentation. Most plants have never plotted it.
A worked example in kVA, with no money in it
Take a plant with a contract demand of 1,500 kVA. Suppose the recorded demand for the month is 1,250 kVA, and suppose the tariff order in force defines billing demand as the higher of recorded demand and a fraction F of contract demand. If F is high enough that F x 1,500 exceeds 1,250, your billing demand is F x 1,500 and your actual restraint that month bought you nothing.
Now suppose one block in the month hit 1,620 kVA. Recorded demand becomes 1,620, which is above contract demand, so the excess is typically billed at a penal rate on the overshoot and the ratchet window now contains 1,620. You pay for that single block in the month it happened and again, through the ratchet, in later months.
Substitute your own contract demand, your own twelve recorded figures and the fraction F written in the order that applies to you. The arithmetic takes ten minutes. The figure it produces is the one worth attacking.
What actually reduces maximum demand
Ranked by how reliably they work, not by how often they are sold.
| Measure | What it changes | Effort | Caution |
|---|---|---|---|
| Staggering large starts and interlocking them | Stops coincident peaks forming in one block | Low — often a control change | Needs discipline at shift change and after a trip |
| Power factor correction to the tariff threshold | Reduces kVA for the same kW | Low to medium | Harmonic-rich plants need detuned reactors, not plain capacitors |
| Soft starters or VFDs on the largest motors | Reduces starting current and smooths acceleration | Medium | A VFD solves starting and control but imports harmonics |
| Automated load management or load shedding | Sheds nominated non-critical loads before the block average crosses a set point | Medium | Only works if non-critical loads are genuinely sheddable |
| Process scheduling | Moves a batch heater or a mill out of the peak block | Low cost, high organisational effort | Production must agree, in writing |
| Battery storage with a demand-targeting control | Supplies the top of the peak from the battery so the meter never sees it | High | Needs accurate metering, fast control and correct sizing in both kW and kWh |
| Increasing contract demand | Avoids the penal excess charge | Administrative | Reduces penalty exposure, raises the fixed floor — this is surrender, not a solution |
Starting method is where plants most often leave demand on the table. If several of your large motors still start direct-on-line, the comparison in DOL vs star-delta vs soft starter vs VFD is the right place to start, because the right answer differs by load type and duty cycle.
What does not reduce maximum demand
Replacing lighting. Trimming compressed air leaks. Efficient motors. All worth doing, and all of them reduce kWh rather than the peak block. A plant can cut consumption measurably and see no movement in its billing demand, because the fifteen minutes that set the demand were never about lighting.
Solar alone also belongs in this list, with a caveat. Behind-the-meter solar reduces imported units during daylight, but your highest block can land on a monsoon afternoon, during an evening shift, or in the half hour after a cloud passes. Demand reduction from solar is real only when it is paired with control that targets the demand peak.
This is the common mistake we see: a plant installs an array, waits for the demand charge to fall, and finds the energy charge dropped while the demand line barely moved.
Where power factor correction stops working
Correction to the tariff threshold is usually the cheapest kVA you will ever buy back. Beyond that, returns fall away quickly, and in harmonic-heavy plants an aggressive capacitor bank creates a new problem.
Plain capacitors in a plant full of VFDs and rectifier loads form a resonant circuit with the transformer impedance. We have been called to plants where capacitor fuses blew repeatedly and APFC contactors were welding, and the cause was not the panel. It was tuning. In our FAT experience an APFC with detuned reactors behaves entirely differently on a harmonic-rich bus, and choosing a 7 per cent or 14 per cent reactor is an engineering decision tied to your dominant harmonic order, not a catalogue preference — normal APFC vs detuned APFC sets out how to choose.
We push back when a client asks us to simply enlarge an existing capacitor bank to fix a demand problem. If the kVAr is already enough to hold power factor near the threshold, more kVAr will not reduce the peak block, and it may make the harmonic situation worse.
How peak shaving with a battery changes the number
A battery sized and controlled for demand management watches the running block average and discharges when the projected average approaches a set point. The meter records the net. The peak the utility sees is the peak you chose.
Two numbers matter and they are independent. The power rating in kW decides how much of the peak you can cut. The energy rating in kWh decides how long you can hold the cut. A peak that is a sharp 10-minute spike needs power and very little energy; a peak that is a two-hour plateau every afternoon needs both, and the energy figure drives the cost.
Control quality decides whether this works at all. The scheme must see the demand block in real time, predict the block average rather than react to it, and act within seconds. That is an energy management system function rather than a battery function, and the range of what such a system can be set to do is covered in the five EMS applications.
The failure mode worth naming: a battery that discharges on the wrong trigger, exhausts its energy early in the block, and leaves the plant exposed for the rest of it. You have then paid for storage and still recorded the peak.
What goes wrong on the electrical side
Four things, in our experience building and commissioning LT boards and metering panels from our Vadodara works.
Measurement first. Demand control is only as good as the instrument that sees it, and we regularly find metering CTs of the wrong class, wrong ratio, or wired with reversed polarity on one phase. A scheme fed by a bad CT will shed the wrong feeder with complete confidence. Our recommendation: verify CT ratio, class and polarity at FAT and again at site, and compare readings against the utility meter for a full month before trusting any automation.
Second, sheddable load has to be physically sheddable. The scheme needs breakers or contactors it can actually operate, with status feedback. Retrofitting that into an old board is often more work than the control system itself, and it is the item that quietly doubles a budget.
Third, the demand set point lives somewhere. If it lives in a relay nobody can access after the commissioning engineer leaves, it will never be revised when the tariff or the load changes. Put it where the plant can see and change it, with the change logged.
Fourth, any peak shaving scheme that includes a DG set has to deal with synchronisation, and that is a protection and control design, not a wiring exercise. Before assuming the DG is the cheap part of the answer, work out your own generation cost using the method in what a diesel generator costs per unit.
What to collect before you commit to any solution
Twelve months of bills with the demand lines extracted onto one sheet. Fifteen-minute interval data for one representative month, which most utilities supply on request. A list of your largest loads with starting method and run pattern. Your shutdown restart procedure. The tariff order clause defining billing demand, printed, with the ratchet clause marked.
With those five items an engineer can tell you in an afternoon whether your demand problem is a control problem, a power factor problem, a scheduling problem or a storage problem. Without them, anyone proposing a solution is guessing. We have seen Gujarat plants spend a year on the wrong one of the four because nobody pulled the interval data first.
The governing principle
Energy charges reward efficiency. Demand charges reward coincidence control. Reducing what you use and reducing when you use it together are two different engineering problems, and the second one is usually where the uncontrolled money is.
Find your worst fifteen minutes of the last twelve months. Then find out what was running during it. Everything else follows from that.
Frequently Asked Questions
How is maximum demand charge calculated?
The meter records the average kVA over each fixed integration block — commonly 15 or 30 minutes — and the highest block average in the month becomes your recorded maximum demand. A billing demand is then derived from that and from your contract demand using the formula in the tariff order in force, and the demand charge is that billing demand multiplied by the rate in the same order. The rate and the formula both change by order, so read the version that applies to your period.
What is the difference between contract demand and maximum demand?
Contract demand is the capacity you have contracted with the utility, fixed by agreement until you apply to change it. Maximum demand is what your plant actually drew in its worst block of the month, which changes every month. You are billed on a billing demand derived from both.
Why am I billed in kVA instead of kW?
Because the network is sized by current, and current follows apparent power rather than real power. kVA equals kW divided by power factor, so a plant with poor power factor presents more kVA to the utility for the same useful work. Correcting power factor reduces the kVA on which the demand charge is computed.
What is a ratchet clause on an electricity bill?
A ratchet or minimum-billing clause sets your billing demand to a stated fraction of the highest demand you recorded within a defined past window, even in months when your actual demand was lower. It means one bad block can raise your bills for months afterwards. The window and the fraction are defined in the tariff order and must be read there.
Does improving power factor reduce maximum demand charge?
It reduces the kVA recorded for the same kW, so yes, up to the point where power factor is near the tariff threshold. Beyond that the returns fall away sharply, and in a plant with significant harmonic content more plain capacitance can cause resonance problems rather than savings. Correct to the threshold, then attack the peak itself.
Will rooftop solar reduce my demand charge?
Not reliably on its own. Solar reduces imported units during daylight, but your highest recorded block may fall on a cloudy day, in an evening shift, or in a monsoon month when generation is low. Dependable demand reduction requires load management or storage with a control scheme that targets the demand block.
How much battery capacity do I need for peak shaving?
It is set by two independent numbers: the kW you need to cut from the peak, and the kWh needed to sustain that cut for the full duration of the peak. A short, sharp spike needs power and little energy; a long afternoon plateau needs both. Both come from your own interval data, not from a rule of thumb.
Can increasing my contract demand reduce my bill?
It can remove the penal excess-demand charge if you are chronically overshooting, but it raises the fixed floor you pay every month and may raise the ratchet base. It treats the symptom. Do the load-coincidence work first, then set contract demand to the level you genuinely need.
What is the integration period and why does it matter?
The integration period is the fixed block over which the meter averages power before storing the value, commonly 15 or 30 minutes. A short inrush barely affects a 15-minute average, while a load running for most of a block affects it substantially. Knowing your period tells you which events you must control and which you can ignore.
Can a diesel generator be used to reduce maximum demand?
Technically yes, by running the DG in parallel during peak blocks, but it requires a synchronising scheme with proper protection and it substitutes an expensive unit cost for a demand charge. Build the DG cost per unit honestly — fuel at the actual specific consumption, plus lubricant, maintenance, manpower and capital recovery — before assuming it is cheaper. For many plants it is not.
Why did my demand charge stay the same after an energy efficiency project?
Because efficiency projects reduce kWh, and the demand charge is set by a single peak block. Lighting upgrades, leak repairs and efficient motors cut consumption without necessarily changing what was running during your worst fifteen minutes. Demand needs coincidence control, not consumption reduction.
What data should I collect before approaching anyone about demand reduction?
Twelve months of bills with contract demand, recorded demand, billing demand and power factor extracted onto one sheet; 15-minute interval data for at least one representative month; a list of large loads with their starting method; your shutdown restart procedure; and the clause in the tariff order that defines billing demand. These five items convert a sales conversation into an engineering one.
Related Guides
- APFC panel sizing and kVAr calculation
- Normal APFC vs detuned APFC
- Load list and transformer/incomer sizing
- Is my factory suitable for solar and battery storage
- The five EMS applications
- Solar and BESS payback and IRR
- Product page: APFC panel
- Service: Battery storage (BESS) EPC
Talk to Us About Your Plant
Wisdom Techno Solutions manufactures the electrical system that sits at the centre of a solar and storage installation — the LT panel, the switchgear, the metering, the protection and the control — at our works in Vadodara, Gujarat. We build APFC panel, and the same engineers who design an assembly test it before it ships and commission it on site.
Wisdom Techno Solutions is a licensed partner for three design-verified switchgear platforms: Rittal Ri4Power (Rittal Solutions Partner since 2021 — the Ri4Power system carries Intertek ASTA design verification to IEC 61439-1 and -2 Edition 3, with copper and aluminium busbar systems up to 6300 A), C&S CX (Licensed Partner since 2023, up to Form 4B Type II) and Siemens SIEPAN Elite/8PU (IEC 61439 Solutions Partner since 2024). We build to each system's verified configuration and routine-verify every assembly in-house at our Vadodara, Gujarat facility, for industrial, EPC and infrastructure projects across India. For a solar or storage plant that covers the LT switchboard, the BESS PCC and the AC and DC distribution.
If you want to know whether your plant is a candidate, send us twelve months of electricity bills and one day's load profile. That is enough to size a system, show where the savings would come from, and tell you honestly if the answer is no.
The quickest route is our free feasibility study: send the bills and we come back with the sizing and the numbers behind it. To work out your own numbers first, use the free demand charge calculator.
Send your single line diagram, your load list or your last twelve bills, and we will come back with the electrical scheme and the numbers behind it.
Incomer metering, demand-side switchgear and capacitor control are built at our LT panel manufacturer in Vadodara.