How We Answer

First, we check the capacity with the utility's data.

Connecting the utility account brings in 12 months of 15-minute demand and the tariff. There is no visit, no photograph and no sensor at this step.

The utility's data gives the first answer.

The highest demand of the year is 220 A on a 400 A service, and the chargers fit. The rest of the work makes that answer one an engineer can sign.

Then we digitize the panel.

A technician photographs the open panel, the schedule and every nameplate, with the panel live. The record is needed for the engineer's signature and for anything below the meter.

Every circuit is written down properly.

Each breaker becomes a typed row with its circuit number, what it serves and its rating.

The panel now has a record.

A QR code goes on the door. Scanning it opens this panel's schedule, its photographs and the calculation drawn from them.

We record what is on every circuit.

Each circuit is traced to the equipment it feeds, and that equipment's nameplate is recorded beside it.

This panel needs no sensors.

Sensors go on only when the utility data cannot settle the question. None of the three reasons applies here, so the calculation stands on the utility's own record.

The existing load is the measured peak.

Where your local code accepts measured demand, the existing load is the highest demand actually recorded over a year: 220 A here. It does not guess at everything that could run at once.

The code adds its margin.

The margin your local code requires goes on the recorded peak. At 125%, the margin in this example, that makes 275 A. The calculation cites the edition the approving authority has adopted.

The chargers are added on top.

4 chargers of 7.7 kW, calculated under the code's own rule for chargers, come to 85 A.

It fits, with 40 A to spare.

360 A against a 400 A rating. The chargers can go in on the service the showroom already has.

A licensed engineer signs it.

The engineer reviews the calculation and signs it under their own licence.

A first estimate says the panel is tight.

The panel already has a record: every breaker scheduled, every load written down, and a QR code on the door. The Determination is a high-level estimate of the capacity the service has, and it says the 400 A service is tight. The next step is to measure it.

We fit sensors to every circuit that matters.

Split-core sensors close around each circuit's conductor, and flexible coils wrap the large conductors of the mains. The panel stays live and nothing is disconnected. Each sensor is labelled against the panel record.

They record through a full operating cycle.

Current and voltage are recorded on every sensor for 30 days, through the facility's busiest days. Then the sensors are collected.

A reading every 15 minutes, on every circuit.

That is the interval your utility bills demand on. The chart shows one week of current on the main service, with its 400 A rating.

The peak crosses the line the code allows.

The code adds a margin to the recorded peak, 125% in this example, so on a 400 A service the peak has to stay under 320 A before anything is added. Every weekday morning it goes over. The highest reading is 336 A, on Tuesday, 7:15 am.

The peak comes from a few circuits at once.

Each band is one circuit on that Tuesday. At 7:15 the rooftop units are starting, the water heater is recovering, the kitchen is on and the parking lights are still lit.

Here is your service.

The same peak, drawn as one bar of amperes. The line is the 400 A rating.

This is what the code counts.

336 A at the peak, and the code's margin goes on top. Counted, that is 420 A, 20 A over the rating before anything is added.

So the new load has nowhere to go.

The EV chargers need 48 A. On this reading there is no room for them, which is why the quote exists.

We look at it circuit by circuit.

Each segment is one circuit's share of the 336 A peak, drawn at its real width.

Life safety, and anything you tell us to leave alone.

These circuits are set aside before anything else is decided, and they are never touched. Here that is life safety, refrigeration and the kitchen.

This much of the peak does not need to be there.

• The parking lights are on day and night. 16 A.
• The water heater recovers during the morning peak. 12 A.
• The supply fan runs at full speed around the clock. 14 A.
• The 3 rooftop units start at the same moment. 34 A.

76 A is recoverable.

The detailed analysis finds the capacity that can be unlocked, circuit by circuit, and leaves the rest of the operation as it is.

Each finding is matched to an intervention.

An intervention is named by what it does to the peak, such as moving a load to another hour or starting loads one at a time. The design says which loads change, what sheds at the peak and how many kW come off. It is fitted to 12 months of the facility's utility demand and 4 to 8 weeks of its key circuits recorded in the peak season.

Simulated, the Tuesday peak comes down by 76 A.

The chart shows the main service on the peak day, as measured. With the design applied, the 7:15 peak falls from 336 A to 260 A. The same chart is drawn again once the panel has been measured after the work.

An electrician fits two timers.

One turns the parking lights on at 7 pm and off at 6:30 am. The other has the water heater recover between 4:30 and 6 am, before the store opens. The work is done under the electrician's own licence.

The controller staggers the rooftop units.

The 3 units used to start together at 6:45 am. The controller Wattif runs on site now starts them one at a time from 4:45, so no two pull their starting current at once. Nothing is replaced.

An electrician fits a drive to the supply fan.

The fan used to run at full speed around the clock. With the drive it slows whenever the space needs less air.

Every change is photographed and checked.

The field app records what was set and photographs what was fitted, against the circuit. We check each one against the specification before the sensors go back in.

We measure again, on the same instruments.

The sensors go back on the same conductors for another window. The result is the facility's own measured peak.

76 amps back.

Measured, the peak falls from 336 A to 260 A. Counted with the code's margin, that is 325 A, and 75 A of the rating is free.

Now it fits.

The EV chargers' 48 A lands inside the 400 A rating, with 27 A to spare.

Two of the changes also cut energy.

Staggering the rooftop units and the water heater timer move energy to another hour. The parking-light schedule and the fan drive remove it. The shaded area is energy no longer used.

63,400 kWh a year less.

The same changes take 25 kW off the peak. Both figures are read off the same instruments as the amperes.

Signed by a licensed engineer.

The panel measured again states the recovered result, and a licensed engineer signs it under their own licence.

The panel's record shows its new capacity.

It lists each change, what it acts on and what settings it carries. The controller stays on site, and Wattif keeps it running.

No upgrade.

The EV chargers go in on the 400 A service you already have.

The capacity report is where this starts.

Recovered, the mall's peak is 236 A. Counted at the code's margin and with the chargers at 80 A, the service reads 375 A against 400: fits, with 25 A to spare. The code allows the 80 A because a listed load management system will hold it.

The panel record names what the controller can reach.

Every circuit the capacity report wrote down: the rooftop units, the water heater, the supply fan, and now the charger feeder. The controller acts only on circuits that are on the record.

Unmanaged, the chargers would push the service to 427 A.

Shoppers plug in through the afternoon, when the mall is already at its peak. Six chargers at full rate on top of 236 A is 427 A on a 400 A service, over the rating for about 3 hours a day.

Held under 80 A, the service peaks at 316 A.

The same afternoon with the chargers sharing 80 A. Each car charges a little more slowly for about 10.8 hours, the service never sees more than 316 A, and the capacity report's count is what the panel actually carries.

The rules the controller runs, written by an engineer.

• Share the current.
• Hold the limit at the peak.
• Stagger the rooftop units.

Ongoing changes run on a controller as fixed rules an engineer wrote and signed. Every action it takes can be traced to one of them. It is not an optimizer and it does not learn.

A licensed electrician fits it. Wattif runs it.

The controller is ours, on the mall's wall, fitted by a licensed electrician alongside the charger feeder. Wattif commissions it, keeps it running and stands behind the limit. Wattif performs no electrical work.

Every month's peak, read on the panel: 316 A at most.

Twelve months of the service's highest 15 minutes with the controller running. The mall's own peak moves with the seasons; the chargers never add more than 80 A to it. The highest month reads 316 A against 400.

The limit held is the result you pay for.

• Each month's peak on the service, measured, against the rating.
• The chargers' share of it, held under the limit.
• The rules as they ran, and any month the controller intervened.

A monthly record signed by a licensed engineer under their own licence. The fee is for the kW kept inside the limit, read on the panel, and for nothing else.

We read 12 months of the utility's interval data.

Connecting the utility account brings in every 15-minute reading of the year, with the tariff. There is no visit at this step.

Each month's charge was set by one interval.

The report finds that interval in every month. The highest of the year is 510 kW, in August, on a Friday at 4:45 pm. Summer months set the largest charges.

The same climb, every afternoon.

The week around August's peak shows one shape each day: demand rises through the afternoon as guests arrive, then eases through the night. Friday, the busiest check-in, is the highest.

It lands at 4:45 pm, as guests check in.

Rooms cooling down, the laundry still drying and the kitchen starting dinner all land in the same quarter hour. That quarter hour is what August is billed on.

The panel record says what could be running.

The capacity report already digitized the hotel's main panel. Every circuit's equipment and nameplate are on record, so the interval is matched to real loads.

Each circuit's share of that interval.

Where circuits have been measured, each share is read directly. Where they have not, it is attributed from the panel record and the equipment's own schedule, and the report says which.

510 kW, split by what drew it.

The same interval as one bar, one segment per circuit, at its real width.

Four things put the peak where it is.

• Every occupied room cools down at once as guests check in. 40 kW.
• The laundry dryers run through the afternoon. 45 kW.
• The kitchen hood fans run at full speed from 4 pm. 12 kW.
• The pool heater runs through the peak. 24 kW.

Together they account for 121 kW of the 510 kW peak, and so 24% of August's demand charge.

The charge is priced on the hotel's own tariff.

This hotel's tariff charges for each kW of the month's peak, so August's charge rises and falls with its 510 kW interval. The report reads each tariff's own terms, including any ratchet, before it prices anything.

Chargers that start at check-in raise the peak.

8 chargers for guests add 62 kW. Plugged in at check-in, they land on the interval that already sets the charge: the peak rises to 572 kW and August's charge by 12%.

The same charging overnight leaves it where it was.

Held until after 10 pm, the same energy stays under today's peak of 510 kW, and the charge stays where it was. The report states both.

What you receive.

• The interval that set each month's charge, with its date and time.
• What set each one, named circuit by circuit.
• What it costs on your tariff, today and with the load you are planning.

It is reviewed by an engineer and not signed, because it explains a bill rather than calculating capacity.

Where the peak is the problem, it can be taken down.

Taking the four causes out of the peak would leave 389 kW. Capacity Recovery goes on to design the changes on the hotel's own demand, and the same panel is measured again once they are in.

Peak Determination named four causes, 121 kW between them.

• Every occupied room cools down at once as guests check in. 40 kW.
• The laundry dryers run through the afternoon. 45 kW.
• The kitchen hood fans run at full speed from 4 pm. 12 kW.
• The pool heater runs through the peak. 24 kW.

That is the starting point. 121 kW of the 510 kW peak is load that need not be in that quarter hour, and taking it out would leave 389 kW.

The panel record says what each change can act on.

The guest room air conditioners, the laundry, the kitchen hood fans and the pool heater are each on circuits the capacity report already wrote down. A change is designed only for a load that is on the record.

Each cause gets a change, designed on the hotel's own demand.

• Guest rooms: Pre-cool the rooms from 2 pm, so check-in cooling is spread over three hours instead of one.
• Laundry: The dryers finish by 1 pm; the afternoon load is folding.
• Kitchen: The hood fans run to the cooking load, on demand control, instead of at full speed from 4 pm.
• Pool and spa: The pool heater is held off from 3 pm to 7 pm; the water holds its temperature.

Nothing is switched off that a guest would notice. The rooms are cool at check-in, the laundry is done, dinner is cooked and the pool is warm. What changes is when each load draws.

On the peak day, the changes leave 389 kW.

The same Friday in August, simulated with the four changes. The afternoon still climbs as guests arrive, but the interval that set the charge is 121 kW lower, and nothing else in the day is higher for it.

The ongoing changes run as rules on a controller.

• Guest rooms.
• Laundry.
• Kitchen.
• Pool and spa.

Each rule is written and signed by an engineer and every action can be traced to one. The controller is ours, on the hotel's wall, fitted by a licensed electrician. Wattif performs no electrical work.

Every month's peak, read on the panel: 389 kW at most.

Twelve months with the rules running, each month's highest 15 minutes measured on the hotel's own panel against the month before the changes. August reads 389 kW where it read 510.

The kW kept out of the peak is the result you pay for.

• Each month's peak, measured, against the month before the changes.
• Which rule held it, and any month it did not.
• The charge on your tariff, before and after.

A monthly record signed by a licensed engineer under their own licence. The fee is for the kW kept out of the peak, read on the panel, and never a share of the bill.

Every board is written down.

Each circuit, what it serves and its rating. One of them, circuit 23, has no label and nothing on the drawings, and it is drawing.

A meter on every board, reading every 15 minutes for a week.

The same interval the utility bills on, but circuit by circuit, so each load is read where it draws. Nothing else changes while the week is read.

10,635 kWh, and the night never drops below 50 kW.

The week on the main board. The working days climb and fall as the floors fill and empty, and every night settles at 50 kW. The weekend is the same 50 kW for 48 hours.

2 am, split by circuit.

The server room needs 9 kW and should be running. The other 41 kW is lighting on empty floors, rooftop units cooling nobody, a supply fan at full speed, car park fans with no cars, and circuit 23.

4,382 kWh a week drawn for nothing.

• Floor lighting: 372 kWh a week.
• Rooftop units: 930 kWh a week.
• Supply fan: 1,093 kWh a week.
• Car park fans: 732 kWh a week.
• Ckt 23, no label: 1,260 kWh a week.

Each one read at its own circuit, so the figure is measured and not estimated. Circuit 23 is traced to an abandoned chiller pump in the car park, running since the chiller was replaced.

The changes: one done once, four that run every day.

• Ckt 23 traced and retired, done once.
• Lights off when closed, every day.
• Setback when closed, every day.
• Fan slowed to the load, every day.
• Car park fans on CO control, every day.

Circuit 23 is disconnected by the office's electrician. The other four are schedules, a setback, a speed limit and an interlock, written as fixed rules on the controller by an engineer. Wattif performs no electrical work.

The same week, read again: 6,253 kWh, and 16 kW at 2 am.

The nights fall to the server room and little else. The working days are the same shape, because nothing anybody needs has changed. The week's peak is down too, from 85 kW to 69 kW.

Each change, read at its own circuit, every month.

• Each circuit before and after, against the week that was read first.
• The rules as they ran, and the nights they held.
• The month's kWh and its peak, on the panel and on the bill.

A monthly report signed by a licensed engineer under their own licence. Where the peak is the larger problem, Peak Determination picks up from the same meters.

First, the boards involved are written down.

Every breaker on the kitchen's board becomes a typed row: its circuit number, what it serves and its rating. The cooler compressor shares its breaker with nothing on paper.

A meter goes on the suspect circuit, reading every minute.

A clamp on the cooler circuit and one on the fan circuit, recording every minute for a week. Nothing else is fitted and the kitchen keeps working.

The week shows 3 trips, each at the 6 am start.

Three cold mornings, three trips, all at 6:02 am. The mild mornings got through. Whatever it is happens at the start of the day and only when it is cold.

The worst morning, minute by minute.

The compressor starts at 6:00 on top of the kitchen's exhaust fans. The circuit sits at 48 A, over its 40 A breaker, and after 3 minutes the breaker goes.

Two loads land together.

The fans draw 14 A. The compressor's start draws 34 A, and runs on at 22 A against 16 A on its nameplate. The breaker was never going to carry the start on top of the fans.

The cause, named with its evidence.

• The start lands on the fans.
• The start runs longer when it is cold.
• The compressor runs above its nameplate.

Matched against the faults we have seen before: a compressor start coinciding with another load, lengthened by a fouled condenser. The report names the circuit, the pattern and the minutes that show it.

The fix, specified for the restaurant's electrician.

• Start the compressor before the fans.
• Clean the condenser coil.
• What the circuit should read after.

Wattif performs no electrical work. The specification says what to change, on which circuit, and what the circuit should read after. The restaurant's own electrician makes the change.

The week after: 30 A at most, and no trips.

The same circuit, read again. The compressor starts alone at 5:15, the fans come on at 6:00, and the circuit never goes over 30 A. Three cold mornings, no trips.

A licensed engineer signs the result.

• The circuits, measured, with what was running when the fault showed.
• The cause, named, with its evidence.
• The fix as specified and as made.
• The circuit after the repair, read for a week.

The report is reviewed and signed by a licensed engineer under their own licence. The meter can stay on: Electrical Watch is the same circuit, kept watched.

The critical boards are written down.

Each board the clinic cannot lose becomes a typed schedule: every circuit, what it serves and its rating. The theatre board is the one this story follows.

Meters go on every critical circuit, and stay.

A clamp on each circuit of the three boards, reading every minute. The watch is the same meter as a fault study, left on, with an engineer behind it.

The supply fan runs at 13.8 A, week after week.

AHU-2 supply fan serves both theatres. Its nameplate says 15 A. For the first weeks it runs a little under that, as a healthy motor does, and the watch has nothing to say.

Then it starts to climb.

From the third week the running current rises, a few tenths of an ampere a week, with no change to the airflow setpoint and no change to the filters. Nobody in the clinic can hear or see anything different yet.

Week 6: over the line, and the pattern is known.

At 16.5 A the fan has crossed 15.8 A, 5% over its nameplate, on a rising trend. An engineer reads the pattern before anybody is told: current climbing against a fixed setpoint is a bearing wearing, and it was seen on this kind of fan before.

One message, with the board, the circuit and the likely cause.

• Theatres and air handling.
• Ckt 4, theatre board, AHU-2 supply fan.
• A worn bearing. The current rises while the airflow setpoint has not changed.
• Change the fan bearing within two weeks, outside theatre hours. No urgency tonight.

That is what reaches the clinic's dispatcher, checked by an engineer first. Not an alarm, and not a dashboard: a diagnosis the FM electrician can act on, with the time it allows.

The bearing is changed on a Saturday, in week 8.

The clinic's FM electrician books the air handler out while the theatres are idle and changes the bearing. Left alone, the current would have reached 20.8 A by week 10, and the fan would have stopped on a Tuesday with a list booked.

Back to 13.9 A, and still watched.

The same circuit, the week after: 13.9 A, under its nameplate again. The watch goes on reading every minute, and the next pattern gets the same treatment.

A signed condition report every quarter.

• Every alert the quarter raised, with its cause and what was done.
• Each watched board's circuits, as measured, against their ratings.
• The faults the library now knows about this facility.

Reviewed and signed by a licensed engineer under their own licence. The record of each board is kept current, so the next question starts from it.

The board is written down, phase by phase.

Every circuit, what it serves, its rating, and which phase it lands on. The single-phase circuits are the ones to watch: they are where a three-phase board goes out of balance.

A power quality meter on the main board for a week.

Current on each phase and on the neutral, voltage on each phase, and the harmonic content of the current, read every minute while the lines run.

Phase C carries 262 A, phase B 198 A.

An imbalance of 14.7%. One phase's bus runs hot, and every three-phase motor on the board, the drives included, runs a little hotter for it.

29.5% of the current is harmonics.

Six drives draw a current that is far from a sine wave. The third harmonic from each phase does not cancel in the neutral; it adds, and the neutral carries 58 A where it should carry almost none.

The board runs 17 degrees hotter than the boards beside it.

The imbalance heats one bus and the harmonics heat the neutral and the transformer. The heat, the drive failures and the nuisance trips are three symptoms of the same two causes.

The cause, named with its evidence.

• Phase C carries 262 A and phase B 198 A.
• 29.5% harmonic distortion in the current.
• The board runs 17 degrees hotter than its neighbours.

The report names the phase, the circuits that unbalanced it, the drives that distort the current, and the readings that show each.

The correction, specified for the plant's electrician.

• Rebalance the phases.
• Filter the harmonics at the drives.
• What the board should read after.

Three single-phase circuits move to phase B, and the drives get line reactors and a filter. Wattif performs no electrical work: the plant's electrician makes the changes to this specification.

Balanced to 0.9%, the neutral at 14 A.

The same board, read again for a week. The phases read 230, 234 and 232 A, distortion is 9.9%, and the board runs within 2 degrees of its neighbours.

A licensed engineer signs the result.

• The board's phases, neutral and harmonics, before and after.
• The circuits moved and the filter fitted, as specified and as made.
• What the drives now see, and what it means for their life.

Reviewed and signed by a licensed engineer under their own licence. The drive section can stay watched, so the next drive to drift is caught before it fails.