Phoenix Journal · Extraction
A commercial kitchen bill that climbs for no obvious reason is rarely a pricing problem - it is usually a story your own meter data is already telling. Here is how to read it and act.
Case postmortem
The bill landed twelve percent higher than the same month last year, and nobody could say why.
Picture a busy site kitchen: same covers, same menu, same opening hours as the year before, yet the electricity bill kept creeping upward. The operator did the sensible things first. They challenged the supplier, checked the standing charge, and asked whether the unit rate had moved. It had, a little, but not enough to explain the gap. So the increase got filed under "energy prices" and forgotten, the way most unexplained cost does.
Three months later a routine visit told the real story. The main extract fan had been running flat out, twenty four hours a day, seven days a week. Someone had put it into manual override during a busy weekend the previous winter and never switched it back. On top of that, the ductwork was heavily grease laden, so the fan was fighting resistance it should never have had to fight. The kitchen closed at eleven at night, but the meter never rested. That was the leak, and it had been hiding in plain sight in data the operator already owned.
Nothing dramatic failed. That is exactly why it ran for months. Three ordinary problems stacked up, and none of them tripped an alarm.
First, the fan never turned down. Modern extraction is meant to modulate with demand, but a system stuck at full speed ignores the single most useful fact about fans: they obey the affinity laws. Power rises with the cube of speed, so a fan running at forty percent of its rated speed draws only about six percent of the energy it uses at full tilt. Run that fan at one hundred percent through the quiet hours and the mid morning lull, and you are paying cube law money for a trickle of cooking.
Second, the ductwork was dirty, and dirt is not free. As grease builds on filters, canopy and duct walls, the system resistance climbs. The motor has to work harder and longer to shift the same volume of air, which quietly lifts the load every single half hour. The link between dirty extraction and higher energy bills is not a marketing line, it is basic fluid dynamics, and it compounds day after day until someone cleans the system back to a bare metal finish.
Third, and most telling, nobody was reading the meter as data. A year of half hourly readings is 17,520 numbers, and every one of them was available in the supplier portal. The overnight profile - the load when the building should be near silent - sat almost as high as the trading day. That flat overnight baseload is the classic fingerprint of something left running, and it was screaming the answer for anyone who plotted it.
A monthly bill gives you one number, and one number can only ever start an argument. Half hourly data gives you a shape, and a shape tells you what is actually happening in the building. When you lay a busy trading day over a closed Sunday and the two lines barely separate, you have found waste without setting foot in the plant room.
Once you can see the overnight profile clearly, the conversation changes. Instead of "the bill went up", you are asking a specific, answerable question: what should be drawing power at two in the morning, and why is the extract fan on that list? Analytics does not replace engineering judgement. It points the engineer at the right half hour so the judgement is not wasted.
It also lets you compare fairly. Two kitchens on paper can be identical and still run wildly different bills, and until you normalise for covers, opening hours and cooking style you cannot tell a genuine problem from a busy week. If you want to benchmark your kitchens' energy use against peers, the half hourly profile is the raw material that makes the comparison honest rather than anecdotal.
Turning a bill into action is a sequence, not a single purchase. Here is the order that works, cheapest and most revealing first.
None of this is exotic. It is the discipline of treating your meter as an instrument rather than an invoice, cleaning to a real standard so the numbers mean something, and letting the fan follow the cooking instead of the clock. The operator in our story did all six over a single quarter. The overnight baseload dropped away, the fan spent most of its life well below full speed, and the mystery twelve percent stopped being a mystery and started being money back in the business.
Questions
A year of half hourly readings gives you 17,520 data points that show the shape of your consumption, not just a monthly total. By overlaying a busy trading day on a closed day you can see whether equipment is running when the building should be quiet. That overnight baseload is usually where the cheapest, fastest savings are found, before you spend a penny on new kit.
Yes, and it is straightforward physics rather than sales talk. Grease on filters, canopy and duct walls raises the system resistance the fan has to overcome, so the motor works harder and longer to move the same air. Cleaning back to a bare metal finish under TR/19 lowers that resistance, which shows up directly in a lower half hourly load.
Demand control varies fan speed to match real cooking activity using sensors for canopy temperature, smoke and steam, or gas use, rather than running flat out all day. Because fan power follows the cube of speed, a fan at 40 percent speed draws only about 6 percent of full power, so the savings are large. Reputable UK systems claim up to 80 percent off ventilation energy and several appear on the government backed Energy Technology List.
Phoenix Duct Clean · by the numbers
The right kit only helps if the system stays clean. Phoenix degreases canopies, filters and ductwork to TR19 Grease - UK-wide, overnight.