Solar Maximum Demand: Why Your kW Charge Stays

A 1,000 kW peak on Medium Voltage General costs roughly RM89,270 a month in capacity and network charges before a single kilowatt-hour is priced. Put a large rooftop array on the same factory and the energy charge falls immediately. The solar maximum demand saving, the kW line, usually does not move at all. That gap between kWh savings and kW savings is the most common error in Malaysian solar business cases, and under TNB's RP4 tariff structure it is an expensive one, because more of the bill now sits on the per-kW charges than it did before.

Why does solar rarely reduce the maximum demand charge?

Because maximum demand is set by a single 30-minute interval, and solar cannot guarantee output in that interval. TNB bills Medium Voltage customers on the highest 30-minute average demand recorded in the month: RM29.43 capacity plus RM59.84 network, RM89.27/kW/month on General, and RM30.19 plus RM66.87, RM97.06/kW/month on Time-of-Use. There are around 1,440 half-hour intervals in a month. Solar has to win all of them. It only has to lose one for the charge to stand.

Energy charges work the opposite way. Every kWh the array produces is a kWh you do not buy, and those savings accumulate across all 1,440 intervals. That is why solar is a genuinely good investment for energy and a weak one for demand. The two lines on the bill are measured differently, so they respond differently.

Your peak and your PV curve are not in the same place

Most Malaysian commercial and industrial peaks land late in the afternoon or in the evening, after PV output has already started to fade. A factory running a second shift, a mall pulling full chiller load into the evening, a hotel stacking laundry, kitchen and cooling load from 6pm: none of these are covered by a rooftop array. Peninsular irradiance is effectively finished by 7pm, and by 5pm a system is often producing a quarter of nameplate or less.

Even when the peak does fall at midday, the array has to be sized against the whole facility load, not against a comfortable fraction of it. A 500 kWp system on a site with a 1,800 kW peak trims perhaps a few hundred kW at noon in clear conditions. That is an illustrative figure, not a promise, and it holds only while the sky cooperates.

One cloudy interval restores the full grid draw

This is the part that kills the business case. A passing cloud can cut array output by 70 to 80 percent within seconds. If that happens while chillers, compressors and production lines are all running, the grid instantly supplies everything solar was supplying, and the meter records that half hour at full facility load.

TNB does not average your month. It bills the worst interval in it. Twenty-nine days of beautiful clipping at noon are worth nothing on the kW line if one Thursday afternoon in the monsoon transition period produced a two-minute downpour during peak production. Any demand strategy that depends on weather is not a demand strategy. It is a hope.

Does Time-of-Use make solar better for maximum demand?

Not really, and this surprises people. Under RP4 the ToU peak window is 2:00pm to 10:00pm on weekdays, with everything else, including weekends and selected public holidays, treated as off-peak. Maximum demand recorded during off-peak hours is not charged on ToU.

Read that carefully. Solar's strongest hours, roughly 10am to 2pm, are almost entirely inside the off-peak window, where your demand was never billed in the first place. Shaving a morning peak with PV on a ToU account saves you nothing on the kW charge, because that peak was already free. Meanwhile the billed window runs to 10pm, and solar covers at most the first third of it, weakly. If anything, ToU widens the mismatch.

There is a second-order effect worth checking too. Solar reduces the real power drawn from the grid but does nothing about the reactive power your motors demand. Grid-side power factor can drift downward after a large array is commissioned, and the surcharge applies below 0.85 for supplies under 132 kV. Sites have been caught adding a power factor penalty while celebrating their kWh savings. Check your capacitor bank settings after commissioning, not six months later.

What actually protects your maximum demand

Three things, in order of cost. First, know where the peak is. Most operators cannot name the day, time or equipment behind last month's billed interval, which makes any countermeasure guesswork. A live max demand KPI, one that tracks the running 30-minute average and warns before the interval closes, changes the conversation from post-mortem to prevention. The mechanics of that, including sequencing, staggering and interlocks, are covered in our guide to cutting TNB maximum demand charges.

Second, stagger and sequence load. Do not let three chillers, the air compressors and the production line ramp together after a break. Simple start delays and soft-start logic on the largest motors often take 5 to 10 percent off the billed peak with no capital spend.

Third, storage. A battery discharges on command regardless of cloud cover, which is precisely the property solar lacks. That determinism is why battery storage does peak shaving that PV cannot, and why the two technologies should be evaluated on separate business cases rather than bundled into one optimistic spreadsheet.

Size the solar case on kWh, not kW

None of this is an argument against rooftop solar. Medium Voltage energy rates under RP4 sit at roughly 29 to 31 sen/kWh, and a self-consumed solar kWh displaces that plus the AFA component on top. For a daytime-heavy load profile the arithmetic is strong. It is simply arithmetic about energy.

So build the model that way. Take your actual half-hourly consumption data, overlay a modelled generation profile, and count only the kWh you genuinely self-consume. Assume zero reduction in the demand charge unless you have measured interval data proving your billed peak sits squarely inside solid production hours, and even then discount it heavily. If you are still comparing procurement routes, the difference between SELCO and the SARES-style programmes matters more to your returns than any assumed kW saving, and measuring what the array actually delivers is what separates a modelled payback from a real one. Confirm current rates against the official myTNB tariff schedule and TNB's commercial and industrial pricing page before signing anything.

If you want a straight answer on what solar would and would not do to your bill, send us a recent TNB invoice and your interval data. We will show you where your billed 30-minute peak actually falls, what the array would cover, and what it would not. Start at request a demo.