Rate design bill impact analysis on every customer’s own meter data — who wins, who loses, before the regulator asks
Winners and losers of every tariff option, in hours.
See one case, screen by screen ↓Every proposed rate design — a time-of-use window, a higher fixed charge, a demand charge, an EV rate, a slab change — has to go to the regulator as bill impacts: how far each customer’s bill moves, by class and segment, and who is hit hardest. Today a rate analyst exports billing determinants from the CIS, draws a sample of a few thousand accounts, builds the tariff in a spreadsheet, prices a “typical customer” at a few usage levels and writes the exhibit by hand. Then the board asks for another option, and the committee sends a data request on a five-business-day clock, and the work starts again.
A sample hides exactly what the regulator wants to know. In a 2,000-account sample of 812,400 customers, about 120 accounts sit above +15% — too few to say whether they are villas with pool pumps, elderly couples on a small apartment or homes with rooftop solar. Moving cost from kWh into a fixed charge is what creates low-use losers, and regulators look for gradualism: the Ontario Energy Board considers mitigation when an application’s total bill impact passes 10%. Rate shock found after filing costs a contested hearing, customer complaints and credibility — and a tariff that under-recovers costs revenue every year it runs.
Estimated split for one residential tariff option, in working days, by hand and with the solution. The target is three options and a data request in one day.
Six agents price every customer on their own meter data, find the tail and draft the exhibit; the head of tariffs decides and signs.
Lumina Utilities is preparing its residential summer tariff reform 2027 for the Tariff Review Committee, due Friday 30 October. The board’s draft, Option B, adds a summer afternoon peak to reward moving air-conditioning load. Here is how the rate team tests it, screen by screen, in the working solution.
Rania opens Option B — summer peak adder: a AED 25 fixed charge, slabs of 19 / 24 / 29 / 35 fils, +32 fils on weekday kWh 13:00–19:00 from June to September and a 9-fil night credit. Run 27-114 has priced all 812,400 residential customers. The curve shows the median bill at −1.7%, 57.7% paying less, 1.2% at +25% or more and class revenue at −0.1%. Five representative homes are pinned where their bills land.
“6.1% at +15% or more · 49,168 customers · see who they are”
The meter data preparer loaded 28.2 billion 15-minute reads for October 2025 to September 2026: 99.2% actual, 0.6% estimated by the meter system, 0.2% short gaps filled. 805,880 accounts are priced on their own reads; 6,520 — new connections, exchanged meters and long communication gaps — on their monthly bills, flagged in every run. Under today’s tariff the model re-creates AED 4,199.9M against AED 4,212.6M billed.
“Year: 4,199.9 against 4,212.6 · −0.30% ✓” — inside the ±0.5% back-cast tolerance
The tail is not an average. 25,856 large villas — 63.1% of the segment — 12,589 villas, 10,619 low-use apartments and 104 homes with rooftop solar. The most affected districts are Falcon Heights (14,895), Date Grove (8,786) and Sandstone Ridge (6,856). Of 22,800 senior-rebate accounts, 3,140 are in the tail.
“18,420 of the tail villas run a pool pump through the peak — the single biggest driver.”
A 7-bed villa with a pool in Falcon Heights, 103,000 kWh a year, matched by 3,910 similar homes and shown with its account masked. Today AED 36,790; under Option B AED 43,726, +18.9%. The summer peak adder adds AED 9,558, the fixed charge AED 180; the lower slabs take off AED 2,060 and the night credit AED 742. August goes from AED 5,456 to 7,675.
“Air conditioning and the pool pump run through the afternoon. 57% of this home’s summer kWh lands in Option B’s peak window, against 17% for a typical townhouse, so the adder outweighs the lower slabs.”
The committee asks how many accounts would see more than 15% under Option B, by dwelling type, and the impact on senior-rebate accounts. It arrived Monday 12 October at 09:40, due Monday 19 October. The filing and request writer runs the cut on run 27-114 and drafts the answer with its table: 49,168 of 812,400 accounts, and of 22,800 senior-rebate accounts a median of +7.9% and 3,140 at 15% or more — mostly because the fixed charge rises from AED 10 to AED 25.
Every number in the answer is linked to its run. Rania approves; Leena Farouk sends.
The mitigation explorer drafts Option C: the summer peak narrowed to 14:00–18:00 at 24 fils, the fixed charge lowered to AED 15 and the slabs 1 fil higher to keep revenue whole. Rania models it on all 812,400 customers as run 27-121, with Option B as an outline beneath. The median is −1.2%, 58.2% pay less, and 739 customers stay at +25% or more.
“Option C cuts the tail from 49,168 to 15,940 customers · Revenue +0.2%, inside tolerance. Senior-rebate median +1.9%.”
Under Option C, large villas are at a median +11.5% and 11,982 of them at +15% or more; villas +2.8%; homes with an EV −15.5%, every one paying less. Senior-rebate accounts fall to a median of +1.9%, 210 at +15% or more. 6,870 villas still run a pool pump through the peak — a pump timer moved to 00:00–06:00 brings most of them under +15%.
Lists go to customer programmes as segments and counts, never names, unless a programme has consent.
Rania recommends Option C for the filing, and the exhibit redrafts itself in six paragraphs — purpose, method, results, customers most affected, alternatives studied and the senior citizen rebate. Each number carries a margin note to where it came from: the back-cast, the distribution of run 27-121, the pinned customers, the options compared and the senior-rebate register.
“Narrowing the summer peak to 14:00–18:00 and lowering the fixed charge to AED 15 reduces that number by 33,228 while keeping revenue within tolerance.”
Omar reviews what he is signing on one screen: 15,940 customers (2.0%) at +15% or more, a median of −1.2%, revenue +0.2% and a back-cast of −0.30%, both inside ±0.5%, every number linked to run 27-121. He signs; Option B stays in the exhibit as the alternative studied. Leena Farouk, regulatory affairs, is next, to file it with the committee.
Agents model, draft and propose. Nothing goes to the committee without the Head of Tariffs’ signature.
The same tariff reform, seen by the four people who own it — what their study looked like, and what it looks like now.
Six agents price every customer on their own meter data, find the tail and draft the exhibit; the head of tariffs decides and signs.
Loads twelve months of 15-minute reads for every account, aligns them to billing cycles, replaces estimated reads with actuals where they arrived later, fills gaps under 4 hours and flags what cannot be trusted.
Reads today’s and the proposed tariff sheets into billing rules — slabs, time windows, adders, credits and fixed charges — and records the sheet version on the run.
Prices every account’s twelve months under today’s tariff and each option — 9.75 million monthly bills for Option B — and back-casts last year’s billed revenue.
Builds the distribution by dwelling type, use band, EV charging, rooftop solar, senior rebate and district, counts the tail and pins representative customers at points on the curve, with their change explained in plain words.
Proposes changes to the window, adder, fixed charge, credit or slabs that shrink the tail while keeping class revenue inside tolerance, and shows each proposal’s tail, median and revenue.
Drafts the bill impact exhibit and each data-request answer from the run, in plain words, naming customers by segment.
Chooses the option and signs. The agents propose; a named person decides.
Ask in plain words — or ask it to try a change or draft an answer. Every reply points to its run.
Who is in the +15% tail under Option B?
49,168 customers (6.1%): 25,856 large villas, 12,589 villas and 10,619 low-use apartments. 18,420 of the villas run a pool pump through the 13:00–19:00 window; the apartments are there because of the AED 25 fixed charge. 3,140 are senior-rebate accounts.
Would a narrower peak and a lower fixed charge shrink the tail?
Yes. Option C narrows the summer peak to 14:00–18:00 at 24 fils and lowers the fixed charge to AED 15, with slabs 1 fil higher to keep revenue whole. The tail falls from 49,168 to 15,940 (2.0%), revenue is +0.2% and the senior-rebate median drops to +1.9%.
Draft the answer to data request DR-14
DR-14 asks for Option B accounts above +15% by dwelling type and the senior-rebate impact. I am running the cut on run 27-114 and drafting the answer with its table — you review it, then Leena sends it.
Can we trust the model enough to file?
Yes, on three checks. The back-cast re-creates last year’s bills at AED 4,199.9M against 4,212.6M billed, −0.30%, inside the ±0.5% tolerance. 99.2% of the 28.2 billion reads are actual. The 6,520 accounts priced on bills move the result by less than 0.05 points.
13 screens from the working solution, on its sample data. Pick one to see it large.
The rate-shock line, the tolerances, the data rules and who signs are settings, versioned and recorded on the audit trail.
| Setting | Default | Choose from |
|---|---|---|
| Rate-shock line | +15% | +5% to +30%, one point at a time — customers at or above it are the tail and named in the exhibit |
| Severe line | +25% | +20% · +25% · +30% — every customer above it is reviewed by customer programmes |
| Revenue tolerance | ±0.5% of the requirement | ±0.25% · ±0.5% · ±1% of the requirement — an option outside it cannot be recommended |
| Modelling year | Oct 2025 to Sep 2026 | Oct 2025 to Sep 2026 · Jan 2025 to Dec 2025 · Oct 2024 to Sep 2025 |
| Minimum days of meter reads | 300 days | 270 · 300 · 330 days — below it, an account is priced on its monthly bills with its segment’s load shape |
| Customer response | None — structural impact | None — structural impact · Price response, elasticity −0.08 · Both, side by side |
| Back-cast tolerance | ±0.5% | ±0.25% · ±0.5% · ±1% |
| Pinned customers shown only when they match | 50 similar homes | 25 · 50 · 100 similar homes, account numbers masked |
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: AED 42M is 1% of the demo utility’s AED 4,212.6M residential class revenue; ≈ 120 accounts is the +15% tail’s 6.1% share of a 2,000-account sample; the split of days is an estimate of team practice · “typical” = published figures (Ontario Energy Board, mitigation plan above a 10% total bill impact, Handbook for Utility Rate Applications). Lumina Utilities, the people, districts, accounts and the Tariff Review Committee named on this page are fictional — characters and sample data in the working solution.
Bill impact analysis shows how a proposed tariff changes customers’ bills — by class and segment, and who is hit hardest — so a regulator can judge a rate design. Bill Impact Analysis does it for every customer on their own meter data, not on a sample, and drafts the filing exhibit and data-request answers. The head of tariffs decides which option is filed.
Six agents work in one run. They load a year of 15-minute reads for every account, read the tariff sheets into billing rules, price every customer under today’s tariff and the option, back-cast last year’s billed revenue, find who is in the tail and pin representative customers on the curve. In the working solution, Option B was priced for 812,400 customers in one run.
It counts every customer at or above the rate-shock line and breaks the tail down by dwelling type, use band, EV charging, rooftop solar, senior rebate and district. Pinned customers explain each end of the curve in plain words — for the large family villa, the summer peak adder adds AED 9,558 and the pool pump runs through the afternoon.
Yes. The mitigation explorer proposes changes to the window, adder, fixed charge, credit or slabs that shrink the tail, and never one outside the revenue tolerance. In the sample case, Option C cut the customers at +15% or more from 49,168 to 15,940 with class revenue at +0.2%, inside ±0.5%.
Yes. Each request is logged with its clock, the cut is run on the same model run, and the filing and request writer drafts the answer and its table with every number linked. The analyst approves and regulatory affairs sends; the agent never sends anything to the committee.
A pinned customer is shown only when at least 50 similar homes share its profile, and account numbers are masked. Lists for customer programmes carry segments and counts, never names, unless a programme has consent.
Yes. The agents model, propose and draft; they never mark an option for filing and never send. The analyst prepares, the Head of Tariffs and Pricing signs the option and the exhibit, and regulatory affairs files. Every run, edit and signature is on the audit trail.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your tariff sheets, meter and billing data, segment definitions and a past filing — and it goes live once every quality gate has passed. We will walk you through it on one of your own tariff options first.
Every data request answered on time, with one number for one thing.
1-min demoRegulatory Precedent SearchEvery filing position backed by how regulators actually ruled, cited and trended.
1-min demoRegulatory Commitment TrackingEvery directive owned and dated within a day of the order, and closed with its proof.
1-min demoWe’ll run Bill Impact Analysis on your own meter data and a tariff option you are studying.