Planned outage management that never approves an outage while its backup feed is out, and tells every customer in time
No outage approved while its backup feed is out, and every customer told in time.
See one case, screen by screen ↓Maintenance, capital projects, new connections, tree cutting and water-main work all ask for outages, 40 to 60 a week. Each one needs a backup feeder to carry the customers outside the section being worked on, must not lean on the same normally-open point as another job, needs its crew free, must stay clear of event embargoes and weather warnings, and must give every affected customer notice inside the rules. Coordinators hold all of that in spreadsheets, a wall planner and the control engineer's memory.
So conflicts surface late. An outage is approved on a feeder whose backup is already out for another job, and the control room finds it the day before. The outage is cancelled, the jointing crew and kit are stood down, and every customer who was written to is written to again. Or it goes ahead on the wrong assumption, and a 412-customer outage becomes a 1,284-customer one, with medically dependent customers on the wider section who were never on the call list.
Estimated split for one coordinator placing 40–60 outage requests a week, in hours, by hand and with the solution.
Six specialist agents check every outage the moment it lands on a date; the outage coordinator approves.
Lumina Utilities is planning the week of Monday 19 October across Harrow Vale, Kingsmere and Saltmarsh primary substations and the Ridgeway water supply zone. It is Tuesday 13 October, 10:20, and next week's letters must post by Wednesday 14 October at 16:00. Here is what happens to one request, screen by screen, in the working solution.
Leila's plan shows 12 outages and 1,980 customers and properties off, with 5 requests to schedule and 1 conflict. A lane across the top carries the Quarry Road closure on Thursday, an amber wind warning on Friday, the cup semi-final and the half marathon. The one red bar is OR-26-0416, the Hollins Park reconductor, which the event and weather watcher has already flagged.
“Event and weather watcher read the amber wind warning for Fri 23: OR-26-0416, overhead work on KM-07, must move”
Tom Brennan's email of 12 Oct 16:41 asks to replace the Lucy-type ring main unit at H47 on Marsh Lane, which failed its partial discharge survey on 2 Oct. The request reader fills the asset, work, dates, times, crew, section off, backup, kit and requester, each cited to the sentence it came from: 08:00–14:00, Ashby Jointing crew B, 412 customers off between RMU-H46 and RMU-H48, back-fed from HV-14 Quarry Road through NOP-1214 at 3.9 MVA.
One assumption flagged: “Back-feeds the rest from Quarry Road” assumes HV-14 is in service on the day.
Leila picks OR-26-0418 from the tray and the HV-12 Marsh Lane row lights up. Monday and Tuesday: Harrow Vale is on one transformer while T2 is overhauled. Wednesday: amber, 4 medically dependent customers and 412 customers off. Thursday: the backup feeder is out. Friday to Sunday: Ashby Jointing crew B is not free.
“Thu 22 Oct — Backup feeder HV-14 Quarry Road is out”
Thursday is one of the two days the crew prefers, so Leila drops it there. The six checks rerun in the side rail and the bar turns red, with a red link to OR-26-0409: Hamid Rezaei's cable diversion for the Kingsway Metro station takes HV-14 out from 07:00 to 17:00. NOP-1214 has nothing to close onto, and the only other tie, HV-13 Orchard Rise, already carries 1.8 MVA of HV-14's load, leaving 0.3 MVA spare against 3.9 MVA needed.
“Can't go on Thu 22 Oct. Backup feeder HV-14 Quarry Road is out. Without it, 1,284 customers would be off for 6 hours instead of 412, and 11 medically dependent instead of 4.”
The backup feeds view shows the spare capacity on every feeder's backup for the week, after every approved outage, from network model HV-NM 2026.10.2. HV-12 Marsh Lane needs 3.9 MVA to HV-14: 4.9 MVA spare on most days, 0 on Thursday. The Harrow Vale one-line diagram for Thursday shows HV-14 out and NOP-1214 open with nothing behind it.
“Thursday: HV-14 is out all day for OR-26-0409 … HV-12 has no backup that day: NOP-1214 has nothing to close onto.”
One click moves it to Wed 21 Oct, the first clear day. HV-14 takes 3.9 MVA through NOP-1214, with 4.9 MVA spare against 4.29 MVA needed with the 10 % margin. Both transformers are in service, no other outage uses the open point, crew B is free and there is no event, closure or warning near Marsh Lane. One amber remains: 4 medically dependent customers — a home ventilator, two oxygen concentrators, home dialysis — and Marsh Lane Primary School, 312 pupils.
“Clear, with actions · 4 medically dependent customers · 412 customers off”
The approval shows the outage, its date and hours, 412 customers off and the backup through NOP-1214, and the open actions handed to the notice writer. It also says what the approval does not do: the switching stays with network control, and nothing reaches a customer until the communications lead releases it.
“Network control writes and runs the switching from this plan. Nothing is sent to customers until Grace Whitlow releases the notices.”
The bar turns green on Wednesday. The notice writer drafts 412 letters, 386 texts and 241 emails to post by Wed 14 Oct 16:00, books 4 priority care calls and generator MG-07 for the home ventilator at 23 Heron Close, and writes a letter to the school. Everything waits for release.
“Notices drafted · waiting for Grace Whitlow to release.”
Grace sees the four medically dependent customers with what each needs power for and the action — generator on site and a call for E. Barlow, a call 48 hours before for the other three. The school gets its own letter and a call: kitchen and heating off 08:00–14:00. Five streets, 412 customers, 386 with a mobile number, and the plain-English letter and text beside them.
Letter: “If you rely on power for medical equipment and have not told us, call us on 0800 555 0142 so we can help.”
On the notices list, OR-26-0418 shows Drafted with its clock at “tomorrow”: letters post by Wed 14 Oct. The other outages of the fortnight sit beside it, sent or drafting when approved, with their channels and calls. Grace releases them; every step, by agent and by person, is already on the append-only decision log.
“Letters post by 16:00 on the post-by day · texts and emails go at 09:00 the next morning · regulatory floor: 2 days”
The same Marsh Lane outage, seen by the five people who request it, plan it, switch it and tell its customers — what their week looked like, and what it looks like now.
Six specialist agents check every outage the moment it lands on a date; the outage coordinator approves.
Reads each outage request — form, email, work order or project programme — into asset, feeder, section to be made dead, preferred dates, hours on site, crew and the switching the requester assumes, citing every field to the sentence it came from.
Works out the customers in the section off and the load to transfer, finds the backup feeder and its normally-open point, subtracts every other outage that day, checks the margin and checks the primary still has both transformers.
Finds any outage on the same feeder in the same hours, any other outage that needs the same open point, and any crew booked elsewhere, naming the outage and the time.
Lists every customer in the section by street, matches them to the priority care register with what each relies on, and lists schools, surgeries, care homes and pumping stations; if the backup is lost, gives the whole-feeder list.
Every morning and on every placement, reads the councils' events lists, highways permits, school calendars and national weather warnings, matches them to the feeders they touch, and applies the event embargo and the wind rule.
Computes the post-by date from the notice rules and drafts the letter, text and email in plain English, builds the call list for medically dependent customers, books a generator for life-sustaining equipment and drafts letters for sensitive sites.
Approves; communications lead releases notices. The agents propose; a named person decides.
Ask about any outage, feeder, customer list or notice — or say what to prepare. Every answer points to its records.
Why can't OR-26-0418 go on Thursday?
On Thursday its backup is gone. The Marsh Lane job back-feeds 872 customers from HV-14 Quarry Road through NOP-1214, a 3.9 MVA transfer. But HV-14 is out all day for OR-26-0409, the Kingsway Metro cable diversion, and HV-13 is already carrying 1.8 MVA of HV-14's load. So all 1,284 customers on HV-12 would be off for 6 hours, with 11 medically dependent instead of 4. The first clear day is Wed 21 Oct; letters must post by Wed 14 Oct 16:00.
Which outages does Friday's wind warning put at risk?
One: OR-26-0416, the overhead reconductor on KM-07 Hollins Park, 318 customers off, Fri 23 Oct 09:00–15:00. Rule 7 moves overhead work off an amber wind day. OR-26-0402 on Station Road is overhead too, but Harrow Vale is outside the warning area. Northfold's crew is free Mon 26 Oct, and the letters would still be in time.
Draft the notices for OR-26-0418
They are drafted: 412 letters, 386 texts and 241 emails for Wed 21 Oct, 4 priority care calls, generator MG-07 for 23 Heron Close and a letter to Marsh Lane Primary School. They wait for Grace Whitlow to release them.
13 screens from the working solution, on its sample data. Pick one to see it large.
What the checks test and who decides are settings, owned by the network operations manager — here, rules version 9, approved on 1 Sep 2026.
| Setting | Default | Choose from |
|---|---|---|
| Backup headroom margin | 10 % | 0–30 %, set with − and + · spare capacity the backup must keep above the transfer |
| Feeder work at a primary on one transformer | Blocked | Blocked · Warned, control approves |
| Shared normally-open points | On | On · Off — two outages may not rely on the same open point on the same day |
| More than 1,000 customers off — second approver | Marcus Hale | Marcus Hale, network operations manager · Owen Castell, control engineer |
| Event venues | On | On · Off — no work on a feeder serving a venue from 4 h before to 2 h after an event |
| Overhead work under an amber wind warning | Moved | Moved · Warned — underground work is not affected |
| School exam days | On | On · Off — from the councils' calendars |
| Notice periods | Homes 5 working days | Businesses 10 working days · medically dependent: letter + call 48 h before · schools, surgeries, care homes 10 working days · water low pressure 2 working days · regulatory floor 2 days |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “target” = the design goal, measured in the live solution · “estimated” = our estimate: 2,400 planned outages a year × 8 % cancelled inside 48 hours × half of them caught at planning × $6,200 per cancellation (a four-person jointing crew for a day plus generator, traffic management and re-notification), demo assumptions that a pilot replaces with your own; the hours split is estimated too · “typical” = published rules: Electricity (Standards of Performance) Regulations 2015, reg. 12 (£35 domestic, £70 non-domestic when planned-work notice is under 2 days); National Energy Retail Rules, Part 7 (4 business days' written notice for registered life-support customers). Context: NERC IRO-017-1 Outage Coordination. Lumina Utilities, its substations, feeders, contractors, events and customers are fictional, and people named on this page are characters in the working solution.
Software that takes planned outage requests from maintenance, projects, connections and water teams, schedules them, checks each against the network and the plan, and tells the affected customers in time. Planned Outage Management puts every outage on one chart by substation and feeder, and six agents check each one the moment it lands on a date; the outage coordinator approves.
The network checker finds the section to be made dead, the load to transfer and the backup feeder with its normally-open point, then subtracts every other outage that day and checks the spare capacity against the transfer plus the margin, 10 % by default. It also checks the primary still has both transformers. If the backup is out, it says which outage takes it and how many customers would then be off.
The clash finder looks for any outage on the same feeder in the same hours, any other outage that needs the same normally-open point on the same day, and any crew booked twice. The event and weather watcher adds road closures, event embargoes, school calendars and weather warnings. A conflict shows as a red bar with the reason and the first clear day.
The customer impact finder matches every customer in the section to the priority care register and says what each relies on — in the sample case a home ventilator, two oxygen concentrators and home dialysis. The notice writer builds the call list for 48 hours before and books a generator for life-sustaining equipment. If the backup is lost, the list widens to the whole feeder, so nobody is missed.
The notice writer computes each outage's post-by date from your notice rules — in the sample, 5 working days for homes and 10 for businesses, with a 2-day regulatory floor — and drafts letters, texts and emails as soon as the outage is approved. Every notice shows its clock on one list, and nothing is sent until the communications lead releases it.
Yes. Water supply zones sit on the same chart as electricity feeders, with their districts and boundary valves. In the sample, a trunk main valve replacement at Beacon Hill re-zones supply from Millbrook and sends bottled water to two home dialysis customers.
Yes. The agents read, check and draft; they never approve, move another outage, write switching, override an embargo or contact a customer. The coordinator approves, outages over 1,000 customers off need the network operations manager too, the control engineer runs the switching, and the communications lead releases the notices.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your network rules, notice rules, letter templates, crew calendars and a sample of past outage requests — and it goes live once every quality gate has passed. We will walk you through it on a week of your own outages first.
We'll run Planned Outage Management on a week of your own outage requests.