What One Power Trip Actually Costs a Process Plant, and How to Put a Number on It
A power trip costs a process plant whatever it costs to get back to saleable output — not the price of the electricity you did not consume. Almost every plant that has never costed a trip gets it wrong in the same direction: they count the dark hours and ignore the three hours of re-stabilisation, the charge that had to be dumped, and the night shift that stood around on full pay.
The number matters because it justifies, or kills, every piece of backup you are being quoted for. A UPS, a second incomer, a battery system, a DG synchronisation scheme — each is a premium against a loss you have never measured. Measure the loss first. What follows is the formula, the components most plants miss, and a worked example in physical units you can price from your own cost sheet.
The cost of a trip is the cost of the restart, not the cost of the dark
A power trip is any supply event that takes the process out of control — a full outage, a breaker trip on the incomer, or a voltage dip deep enough to drop contactors and stall drives. The process does not care which one it was. It cares that agitators stopped, that an exotherm lost its cooling, that a spinning line snapped, that a forehearth started cooling.
That is why dark hours are the wrong unit of account. A dip of 140 milliseconds can leave the supply restored before the operator looks up and still cost a four-hour recovery, because every contactor in the MCC dropped out and every VFD went to undervoltage trip. The supply came back in a fraction of a second. The process did not.
Our recommendation to every plant head who asks us to quote changeover equipment is the same: before you compare schemes, write down how many hours it takes you to get from "power is back" to "product is in spec". That figure drives everything downstream.
The formula: six terms and one multiplier
C = B + (T × M) + S + L + O + K
Every term is in your own money, and every term has a source document somewhere in the plant.
| Term | What it is | Where the number comes from |
|---|---|---|
| B | Batch, melt or charge value lost beyond recovery | Batch card, charge sheet, BOM cost roll-up |
| T | Hours from supply restored to saleable output | Production log timestamps, DCS trend, shift report |
| M | Contribution margin lost per hour of lost output | Costing sheet, per hour of rated line output |
| S | Scrap, rework and off-spec product | QC rejection record, rework hours booked |
| L | Labour paid and not producing | Attendance plus contractor hours, affected shift |
| O | Fixed overhead absorbed over output you did not make | Annual fixed cost ÷ planned annual output hours |
| K | Expedited despatch, delivery penalty, utility restart, maintenance pulled forward | Logistics debits, contract clauses, work orders |
Contribution margin, not selling price, is the right multiplier — you never bought the raw material for product you never made, so full revenue overstates the loss. And overhead is only a real loss if the output is genuinely lost rather than deferred: a plant at sixty percent utilisation can usually make the tonnes up next week, a plant running flat out cannot.
Term one: the batch, melt or charge you cannot recover
Some processes lose everything in the vessel the moment agitation or temperature control stops. Pharmaceutical and specialty chemical batches under a validated recipe are the clearest case — the batch is typically written off, with a deviation report behind it. A foundry loses a melt if the holding furnace cools past the point where the charge can be brought back without quality risk. A glass line does not lose a batch; it risks losing the tank.
Write B as zero where it genuinely is zero. A rolling mill or an assembly line rarely scraps work in progress on a trip; it pays in T and L instead.
Term two: restart and re-stabilisation time
This term dominates most honest calculations, and nobody logs it. Re-stabilisation time is the interval between electrical restoration and the first unit of product that passes QC. It covers the sequence walked by hand: boiler light-off and steam pressure, chillers back to setpoint, vacuum pulled again, compressed air recovered, extruder barrels back to profile, a line re-threaded, a dryer brought up without thermally shocking the product.
Pull three months of production logs and look at the first saleable-output timestamp after each unplanned stop. Many plants find a nominal twenty-minute outage consistently produces a five or six hour recovery, and that the variance is driven by which utilities had to be restarted cold, not by outage length.
In our site commissioning work the same ordering problem appears again and again: the restart sequence lives in one engineer's head, and when that engineer is on leave, recovery takes twice as long. We push back on this at handover. Write the sequence down, with interlocks and prerequisites named, and keep it in the control room — it is the cheapest reduction in T you will ever make.
Terms three and four: scrap, rework, and labour that is paid and not producing
Scrap is usually under-counted, because it is recorded against quality rather than against the power event. Build the link: when QC rejects a lot, the log should carry a reason code that returns it to the trip. The same applies to rework hours, product purged to drain, and the first run after restart that is deliberately made to a lower grade.
Labour is simpler. Count everyone on site for the affected shift — payroll and contract — who was paid and not producing, including the maintenance team called in. L is usually larger than expected, because contract packing and handling crews are paid per shift and cannot be sent home at short notice.
One judgment from the shop floor: when a plant tells us "the operators used the time for cleaning", we accept that only if the cleaning was on the plan and is now not repeated. Work moved around is not work saved.
Term five: fixed overhead carried by output you did not make
Absorbed overhead is the share of your fixed annual cost that every planned production hour is supposed to carry. Depreciation, rent, insurance, salaried staff, security, licences — divide the annual total by planned annual production hours.
This is where the argument between the plant head and finance starts, so settle it with one question: is the order book full? If yes, the lost hours are lost output and O is real. If no, O is a timing effect and should be excluded. Do not let a backup vendor include it on your behalf.
Term six: the knock-on costs most plants forget
K is where the lumpy numbers hide. Expedited despatch to recover a shipment date. Contractual delivery penalties. Demurrage on a container booked and missed. Effluent treatment out of specification because the blowers stopped, which can bring a separate regulatory consequence.
There is an electrical knock-on too. A hard trip and a hard restart are themselves events — inrush on every transformer and motor at once, and breaker operations counted against mechanical life. We have opened more than one incomer panel after a trip-heavy monsoon and found heat-damaged joints nobody had reason to look for, which argues for a thermographic survey after a severe event rather than only on the annual plan.
A worked example in physical units, which you then price yourself
This example is illustrative and every assumption is stated. Substitute your own figures before you believe the answer.
Assume a continuous polymer compounding line rated at 1.2 tonnes per hour, running three shifts. Assume a voltage dip at 02:10 drops the MCC contactors. Assume, from the plant's own log, that first in-spec output comes at 08:00. Assume the extruder purge produces off-spec material at a known rate during start-up. Assume 16 people on site for the affected shift, payroll and contract combined.
| Line | Assumption used | Physical result |
|---|---|---|
| Lost production hours (T) | 02:10 dip, 08:00 first in-spec output | 5.8 hours |
| Lost saleable output | 1.2 t/h × 5.8 h | approximately 7.0 tonnes |
| Purge and off-spec on restart | stated purge rate over the ramp | approximately 0.9 tonnes |
| Idle labour | 16 people × 5.8 hours | approximately 93 person-hours |
| Consumables pulled forward | one screen pack and filter change | one change |
| Knock-on | one truck re-booked for next day | one re-booking |
Now price those five lines from your own cost sheet — contribution margin per tonne, off-spec recovery value, fully-loaded labour cost per person-hour, consumable cost, logistics debit. That is C for one event. Multiply by the number of comparable events in the last twelve months and you have an annual exposure figure, which is the only number that makes a backup case arguable either way.
The example claims no typical cost for polymer work, pharma, textiles or foundry, because there is none. Two plants in the same industry on the same estate differ by a large factor, because their restart sequences differ.
Not all trips cost the same: a severity matrix
| Event | Electrical character | What usually survives | What drives the cost |
|---|---|---|---|
| Dip under ~200 ms | Deep sag from a remote network fault | UPS loads, drives with ride-through set | Contactor drop-out across the MCC |
| Interruption, seconds to a minute | Auto-reclose, or upstream trip and restore | UPS loads only | Process state lost before the DG picks up |
| Sustained outage | Feeder or transformer fault | Whatever the DG is sized to carry | Whether essential loads were correctly identified at design stage |
| Internal trip | Earth fault, overload, mis-coordination | Depends entirely on discrimination | Whole plant drops when one feeder faulted |
| Partial trip on one bus | Incomer or bus section fault | The healthy bus section | Whether the coupler scheme can transfer load safely |
The fourth row is worth staring at. A trip that takes the whole plant out because an upstream breaker operated before a downstream one is entirely within your control, and the fix is a study rather than a capital purchase. If you have never had a selectivity and discrimination review done, that is where the cheapest reduction in trip frequency usually sits.
Why a 150-millisecond dip can cost as much as a four-hour outage
Because the process measures its own state, not the supply. Industrial contactors generally drop out when coil voltage falls below roughly 70 to 75 percent of rated for more than about half a cycle to a few cycles, and a VFD trips on DC bus undervoltage in a comparable window unless ride-through has been configured. So a dip the utility would not call an interruption stops every contactor-held motor at once.
| Measure | What it protects | Honest limitation |
|---|---|---|
| Control supply from a small UPS | Contactor coils, PLC, relays, HMI | Keeps nothing turning; prevents spurious drop-out |
| VFD ride-through / kinetic buffering | Drives on inertial loads | Short dips only |
| Undervoltage relay with a deliberate delay | Stops nuisance tripping on shallow dips | Must be coordinated, not just delayed |
| Interlocked auto-restart in the PLC | Reduces T by sequencing the restart | Dangerous if written without process interlocks |
| UPS on critical process loads | Keeps the loads that define the batch alive | Costly per kW; needs a defined essential load list |
| DG with automatic start and changeover | Sustained outages | Start-to-load time means the dip is already lost |
The order matters. Our view, formed on site and on the test floor at our works in Vadodara rather than from catalogues, is that the control-supply UPS and the drive settings come first in almost every plant, because they address the most frequent event class at the lowest cost. A DG, however well engineered, cannot save you from a 150-millisecond dip. We make that point early in every enquiry, because a plant that buys a generator expecting dip immunity has bought the right machine for the wrong problem. A DG synchronisation and load sharing scheme answers the sustained-outage class, and answers it well.
What the electrical system can prevent, and what it cannot
It can prevent nuisance trips from poor coordination, whole-plant loss from a single feeder fault, control-circuit drop-out on shallow dips, long changeover gaps, and the second trip caused by a badly sequenced restart. It cannot prevent a fault on the utility network, a dip caused by someone else's motor starting two kilometres away, or the physics of a process that needs three hours to come back to temperature.
At one Gujarat plant where we replaced an ageing incomer board, the plant's own records showed eleven unplanned stops in a year, and the coordination review found six of them were internal — a downstream earth fault that should have cleared locally was taking out the incomer. That is a design defect, not bad luck, and we said so plainly in the report. Our retrofit and system upgrade work exists largely because installations drift out of coordination as loads are added over fifteen or twenty years and nobody re-runs the study.
How to start measuring this month without spending anything
- Add two fields to the shift report: time supply lost, time first in-spec output resumed. After ninety days you have a real distribution for T.
- Tag QC rejections and rework hours with a power-event reason code, so S stops hiding inside quality statistics.
- Ask finance for one figure — contribution margin per hour of rated line output. Not revenue. Margin.
- Ask your panel builder or contractor for the incomer relay event log and the breaker operation counters. There are usually more events than the production log shows, because brief ones never get reported.
With those four in place the annual exposure builds itself, and when a vendor arrives you can do the only sum that matters: annual exposure against annual cost of ownership, for the event class that the proposal actually addresses. The same discipline underpins any energy investment, which is why we recommend it before assessing whether a factory is a candidate for solar and storage.
Common mistakes when putting a number on a trip
- Counting only the hours the plant was dark, and leaving out restart and re-stabilisation time, which is usually longer than the outage itself.
- Averaging every trip together. A 150-millisecond dip in the middle of a critical batch and a planned shutdown are different events and belong in different rows of the severity matrix.
- Pricing lost output at its sales value instead of its contribution margin, which overstates the cost, or at raw material alone, which understates it.
- Forgetting the knock-on costs: catch-up overtime, expedited freight, a missed despatch, and the wear a hard stop puts on drives and gearboxes.
- Working from memory instead of a trip log. Without date, time, duration and what stopped, the number will not survive a review by the finance team.
- Assuming storage or a UPS removes every trip. It can carry the critical bus through a dip or a short outage; it does not fix a fault inside the plant, poor protection coordination or a weak incomer.
The principle to hold on to
Count the recovery, not the outage. The electricity you did not buy during a trip is trivially small; the hours spent getting back to saleable product, the batch written off and the shift you paid for are not.
One last thought. The plants that recover fastest are not those with the most backup equipment. They are the ones where the restart sequence is written down, the essential load list is current and the protection is genuinely coordinated — three things that cost engineering time rather than capital.
Frequently Asked Questions
How do I calculate the cost of one power trip at my plant?
Add six terms: batch or charge value lost beyond recovery, restart hours multiplied by contribution margin per hour of lost output, scrap and rework, idle labour for the affected shift, absorbed fixed overhead where the output is genuinely lost, and knock-on costs such as delivery penalties. The restart-hours term is usually the largest and the least well recorded.
Why is restart time more expensive than the outage itself?
Restoring the supply is not the same as restoring the process. Utilities have to be brought back in sequence — steam, chilled water, vacuum, compressed air — and temperature-controlled equipment driven back to profile before product meets specification. A twenty-minute outage commonly produces several hours of re-stabilisation, and that interval is where the money goes.
Should I use selling price or contribution margin for lost output?
Contribution margin, which is selling price less the variable cost you did not incur. You never bought the raw material for product you never made, so using full selling price overstates the loss, sometimes by a large factor. Finance can give you this per hour of rated line output in a few minutes.
Does a voltage dip really cost as much as a full outage?
It can. Industrial contactors typically drop out when coil voltage falls below roughly 70 to 75 percent of rated for more than a few cycles, and drives trip on DC bus undervoltage in a comparable window. A dip of a tenth of a second therefore stops every contactor-held motor at once, and the recovery that follows is the same recovery you would run after a long outage.
Will a diesel generator protect my process from voltage dips?
No. Even a well-engineered automatic mains failure scheme needs time to detect the failure, start the set, build voltage and frequency and transfer load, by which point a dip lasting a fraction of a second has already stopped the process. A generator is the right answer for sustained outages; dips are addressed by control-supply backup, drive ride-through settings and correctly set undervoltage protection.
What is the cheapest way to reduce trip losses?
In most plants, two low-capital measures: a small UPS feeding the control circuits, PLC and HMI so contactors do not drop on shallow dips, and correct ride-through configuration on the variable frequency drives. After that, a protection coordination review, because a surprising share of unplanned stops are internal faults that should have cleared downstream.
Should absorbed overhead be included in the cost of a trip?
Only if the lost output cannot be recovered. If the order book is full and the hours are gone, the overhead those hours were meant to carry is a real loss and belongs in the calculation. If the plant runs below capacity and the tonnes can be made up later, the effect is a timing shift and should be excluded or heavily discounted.
How many trips does a typical industrial plant have in a year?
There is no reliable typical figure, and anyone quoting one for your industry is guessing. Your own incomer relay event log and breaker operation counters give the real count, and it is usually higher than the production log suggests because short events are rarely reported. Ninety days of logging gives a usable estimate.
How do I tell whether a trip came from the grid or from inside my plant?
Read the protection event record at the incomer and at the downstream feeders, including relay timestamps and fault currents. If a downstream earth fault or overload registered at the same instant, the event was internal and is a coordination problem. If only the incomer saw a disturbance with no corresponding internal fault, the cause was almost certainly upstream.
What is re-stabilisation time, exactly?
It is the interval between electrical restoration and the first unit of product that passes quality control. It is not the time to restart the motors, and it is not the time to reach rated throughput. Defining it as time to saleable output is what makes the cost calculation honest.
How do I use the trip cost number in a backup investment decision?
Build annual exposure: cost per event multiplied by the number of comparable events in the last twelve months, split by event class. Compare that with the annual cost of ownership of the proposed fix, for the specific event class that fix addresses. A solution covering only sustained outages should not be credited with avoiding dip losses.
What should I log from today to make this calculation possible?
Two timestamps on every shift report — supply lost, and first in-spec output resumed. A reason code linking quality rejections and rework hours to power events. One figure from finance, contribution margin per hour of rated output. And a monthly download of the incomer relay event log. Those four inputs are enough to build the whole calculation without new instrumentation.
Related Guides
- AMF vs ATS vs DG synchronization
- 3-pole vs 4-pole ATS
- Breaker selectivity, discrimination and cascading
- BESS with an existing DG and synchronisation panel
- Is my factory suitable for solar and battery storage
- EMS vs BMS vs PPC vs SCADA
- Product page: AMF 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 AMF 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 BESS sizing 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.
Changeover, backup and essential-load distribution are engineered at our LT panel manufacturer in Vadodara.