botminds.aiRequest a demo
SystemSupply Chain, Procurement & Finance

Capital Portfolio Optimization

Utility capital planning that funds the projects removing the most risk, every number traced to its asset records

More risk removed for the same capital, every project traced to its asset data.

See the demoBook a demo
See one case, screen by screen ↓
demo$412.0Ma yearof monetised risk removed by the $1.2B board plan — 61 of 214 projects funded on the efficient frontier
demo39%more risk removed for the same $1.2B than the FY2026 method, ranked by asset age ($296.4M a year)
demo$28.5Ma yearof risk a $1.0B envelope gives up — four named projects, $194.0M of capital, shown before anyone offers the cut
target100%of projects in the filing traced to their asset health and failure records
The problem

Why a capital plan is hard to defend in a rate case

Every cycle a capital planning team has to turn hundreds of requests — replacements, load growth, resilience, data-centre interconnections, water mains and treatment — into one plan that fits the envelope the board and the regulator will allow. The business cases arrive in different years' dollars, some without contingency, some counting the same benefit twice. Each engineering team rates its own risk high, medium or low in its own way, so the ranking follows the oldest asset or the loudest request.

Then finance asks what a smaller envelope would mean, and the spreadsheet is rebuilt over weeks; the answer is a list of projects, not the risk left on the network. When the regulator's staff ask why one project is above the line and the next is below it, the reason is written by hand from memory. Capital enters the rate base only when it is shown to be prudent, and a cost that cannot be shown that way can be disallowed.

typical$208Bof capital US investor-owned utilities planned for 2025 (EEI)
typical$1.1Tof capital planned by the same utilities for 2025–2029 (EEI)
typical$29Bof rate increases requested or secured in the first half of 2025, a record (Utility Dive via POWER)
Where a capital plan's working days goestimated
By hand75 days
With the solution19 days
  • Reading business cases and putting them in one year's dollars15 → 3 d
  • Scoring each project's risk with the engineering teams20 → 5 d
  • Ranking against the envelope, crews and dependencies10 → 2 d
  • Re-running the plan for a smaller envelope15 → 1 d
  • Writing the investment justification10 → 4 d
  • Review and sign-off5 → 4 d

Estimated split for one planning cycle of about 200 business cases, in working days, by hand and with the solution. The sample plan in the demo is the fictional Lumina Utilities' 2027–2029 plan.

How it works

How a capital plan is built

Five specialist agents read, value and rank every project on the efficient frontier; the head of capital planning signs the portfolio.

What comes in
Requests in214 business cases · cases, asset health, failure records, envelope
Agents at work
Business case reader2027 dollars · checks
Then
Risk valuerhealth × failures × values
Then
Portfolio optimiserfrontier + funding line
Scenario analyst
Then
Justification writerevery figure cited
A person decides
Head of Capital Planningdecides and signs
What comes out
Ranked portfolio
Scenarios with bill impact
Regulator justification
One case, step by step

One capital plan, from 214 requests to a signed recommendation

Lumina Utilities runs an electricity and water network. Its 2027–2029 plan has 214 candidate projects asking for $2,310.0M against a $1.2B board envelope; the board paper is due Oct 29 and the rate filing Dec 1. On Tuesday Oct 13 the CFO asks what $1.0B would give up. Here is what happens next, screen by screen, in the working solution.

  1. 01Tuesday morning

    The plan opens on the efficient frontier

    Hana Okoye · Head of Capital Planning

    Every project is a bead on one curve: risk removed a year against cumulative capital, in order of risk removed per dollar. At the $1.2B budget line, 61 of 214 projects are funded for $1,190.0M and remove $412.0M a year of monetised risk; $10.0M is held as contingency and the household bill rises $3.40 a month. The six must-fund projects, $186M, sit hatched at the start. The FY2026 method, ranked by age, sits well below the curve at $296.4M.

    On the frontier: “+$115.6M a year more risk removed than the FY2026 method for the same $1.2B (+39%).”

  2. 02The run

    214 cases read, valued and ranked under 7 constraints

    Business case reader, Risk valuer and Portfolio optimiser

    The run log shows each step. The Business case reader read 214 cases in 4.1 minutes and put every cost in 2027 dollars. The Risk valuer valued every project from 486,200 assets, 11,842 failure records and value framework v3. The Portfolio optimiser funded the six must-fund projects first, then ranked the rest within the mains crews' 42 km a year, two cable jointing teams, a 30% water floor and four dependencies.

    Portfolio optimiser: “Ranked 214 projects under 7 constraints · 0.6 s” — it never drops a must-fund project and never breaks a dependency.

  3. 03Before ranking

    22 business cases corrected, every original kept

    Business case reader

    The Case corrected filter lists the 22 cases the reader fixed: costs priced in 2024 dollars and escalated, benefits counted twice and removed, missing contingency with 10% added, scope that overlaps another candidate. Each correction sits beside the engineer's original figure. The reader flags; it never rejects a case.

    “Shows the original beside every correction”

  4. 04At the line

    Harrow Bay: a risk value built from its own records

    Risk valuer

    Rank 60, the Harrow Bay 132 kV oil-filled cable replacement, $68.0M. Circuits HB-132-C1 and C2 were laid in 1971 and score health 28 and 31. The class curve gives 0.14 major failures a year; Lumina's own record of 3 oil leaks and 1 joint fault in 24 months raises it to 0.19. Times a $54.0M consequence — 38,400 customers off supply at $62.50 a customer-hour, oil into the Harrow estuary, repair, a reportable release, the public near the joint bays — that is $10.3M a year of risk, $0.4M after replacement: $9.8M removed.

    Risk valuer: never invents a failure rate — it uses the class curve when an asset's history is thin, and says so.

  5. 05The evidence

    Five records behind the number, and two fixes to the case

    Risk valuer and Business case reader

    Each record is linked: oil leaks on Mar 4, 2025 (2,900 L), Nov 19, 2025 (3,600 L) and Jun 2, 2026 (4,700 L); a joint fault on Aug 14, 2025 that took circuit 1 out for 31 hours with 21,300 customers held on backfeed; a 2024 condition survey where sheath tests failed on 4 of 22 sections. The case was priced in 2024 dollars, so $6.3M of escalation is included, and an oil top-up saving counted twice is now counted once.

    Valuer run: “$9.8M a year removed · before $10.3M · after $0.4M · every input cited”

  6. 06The CFO's question

    Move the line to $1.0B and see what is given up

    Hana Okoye · Head of Capital Planning

    Marcus Feld asks what the $1.0B in the regulator staff's opening testimony would mean. Hana moves the budget line. Four projects drop: Westmere cast-iron main renewal phase 3 ($52.0M, $7.9M a year), Cobalt Ridge substation flood defence ($41.0M, $6.2M), the Harrow Bay cable ($68.0M, $9.8M) and the Marlow Vale reservoir outlet tower ($33.0M, $4.6M). 57 projects stay funded for $996.0M and remove $383.5M a year; the bill rise falls to $2.83 a month.

    “−$194.0M capital · −$28.5M a year of risk removed (6.9%)”

  7. 07Below the line

    Westmere: 61 bursts in five years

    Sana Iqbal · Water Network Planning

    The Westmere cast-iron mains, laid in 1934 and 1936, have burst 61 times since 2021, and the count rises every year: 9, 10, 12, 14, 16. The fitted trend expects 12.2 bursts next year at $0.69M each, so renewing 14.2 km with PE pipe removes $7.9M of its $8.4M a year. The records include discoloured-water notice DWN-2026-018, with an action plan due March 2027. At $1.0B it drops below the line; it rejoins at $1.05B.

    Asked to keep Westmere at $1.0B, the assistant proposes deferring Pinehurst depot fleet charging and Ashby Vale 11 kV switchgear: $999.5M, and $383.0M a year removed — $0.5M less.

  8. 08Side by side

    Three envelopes, with the risk each one leaves

    Scenario analyst

    The board plan at $1.2B removes $412.0M a year. The CFO test at $1.1B removes $397.6M and drops Harrow Bay and Marlow Vale. Staff testimony at $1.0B removes $383.5M and drops all four. Risk removed is broken out by family — customers and supply 233.2, 228.4 and 220.7 $M a year; water quality 19.0, 19.0 and 16.9 — with the bill impact under each.

    Scenario analyst: “Compares; does not choose.”

  9. 096 seconds later

    The justification, redrafted for $1.0B

    Justification writer

    The writer redrafts seven sections with nine sources. The new section names the four projects that drop, the $194.0M and the $28.5M a year of risk they carry, and why Harrow Bay matters most: three oil leaks and a joint fault in 24 months put its chance of a major failure at 19% a year. The bill section compares $2.83 a month with $3.40 for the board plan.

    Owen Hale's review note: “Staff will ask what a lower envelope gives up. Keep the $1.0B section ready, with the Harrow Bay paragraph.”

  10. 10Signed · Oct 13, 10:12

    Hana signs, with the meaning of her signature

    Hana Okoye · Head of Capital Planning

    Hana makes the $1.0B scenario the recommendation and signs it with the justification drafted for the same envelope. The signature records its meaning — “I recommend this portfolio to the CFO and the board” — and what comes next: Marcus Feld approves the envelope, and Owen Hale files. The signing is recorded on the audit trail and Marcus is notified.

    If the draft and the recommendation are for different envelopes, the solution asks her to redraft first and will not sign.

Who it’s for

Built for everyone who answers for the capital plan.

The same plan, seen by the five people who build, fund, value, defend and argue for it — what their cycle looked like, and what it looks like now.

HO
Hana OkoyeHead of Capital Planning, Lumina Utilities · signs the recommended portfolio
Head of Capital Planning
Before
Owns the 2027–2029 plan: 214 requests, one envelope, a board paper on Oct 29 and the rate filing on Dec 1. “I can rank them. What I can't do is show the regulator why number 58 is below the line and number 57 is above it.”
Now
Moves one budget line and sees which projects make the cut, the risk left on the network and the bill impact, with every project's risk traced to its records.
MF
Marcus FeldChief Financial Officer · approves the envelope
CFO
Before
Sets the envelope and has to know what a cut costs before he offers it in settlement. “If staff push us to a billion, tell me exactly what we stop doing.”
Now
Sees the board plan, his $1.1B test and the $1.0B staff testimony side by side: the projects that differ, the risk removed by family and the bill for an average household.
IA
Dr. Ines AlbrightDirector of Asset Management · owns the asset register and the value framework
Director of Asset Management
Before
Owns the asset register, health scores and failure records. “Every risk number should come from our data, not from whoever shouts loudest.”
Now
Approves the value framework every project is priced with, and sees the failure curve for each asset class fitted on Lumina's own records.
OH
Owen HaleRegulatory Affairs Manager · writes the capital testimony and files
Regulatory Affairs Manager
Before
Writes the capital testimony and answers data requests. “Every project in the filing needs a reason I can cite.”
Now
Gets an investment justification with a source on every figure, redrafted for whichever envelope the regulator's staff put on the table.
SI
Sana IqbalWater Network Planning Manager · puts forward the water mains and treatment programme
Water Network Planning Manager
Before
Puts forward the water mains and treatment programme. “Westmere has had 61 bursts in five years. I need it to be seen.”
Now
Sees Westmere's bursts, its discoloured-water notice and its risk value on the same scale as every grid project, and what it would take to keep it in.
Built on the engine

5 agents. Each with one job, and hard limits.

Five specialist agents read, value and rank every project on the efficient frontier; the head of capital planning signs the portfolio.

Business case reader

Reads each business case, PDF or workbook, and pulls cost by year, contingency, scope, the assets it touches and the benefits it claims. Puts every cost in 2027 dollars with the finance escalation index and flags missing contingency, benefits counted twice and scope that overlaps another candidate.

  • Shows the original beside every correction
  • Flags; never rejects a case
  • Read-only on the business cases
Risk valuer

Links each project to its assets in the register, reads ten years of health scores and failure records, estimates failures a year from the asset-class curve adjusted by the asset's own record, prices each consequence with the value framework and computes the risk before and after the work.

  • Never invents a failure rate
  • Says when the class curve was used — an asset needs at least 3 records of its own
  • Every input cited
Portfolio optimiser

Funds must-fund projects first, then ranks the rest by risk removed per dollar within deliverability caps, programme floors and dependencies, and sets the funding line for any envelope. Reports what each change in envelope adds or gives up.

  • Never drops a must-fund project
  • Never breaks a dependency
  • Does not silently fill the remainder — it is held as contingency
Scenario analyst

Re-runs the portfolio for each envelope or variant and reports the funded projects, cost, risk removed by family, residual risk and the bill impact for an average household, naming every project that differs from the board plan.

  • Compares; does not choose
Justification writer

Drafts the investment justification for the recommended scenario — how projects were ranked, must-fund obligations, what the envelope buys, what a lower envelope gives up and the bill impact — and answers to the regulator's questions, citing every figure to its business case, record or framework.

  • Draft until signed
  • Never files with the regulator
  • No figure without a source
Head of Capital Planning

Signs the portfolio; the CFO approves the envelope. The agents propose; a named person decides.

Ask in plain words

Ask about any project or envelope, in plain words

Ask about any project, the funding line or a scenario — or say what to change. Every answer points to its records.

What do we give up at $1.0B?

Four projects drop and $28.5M a year of risk stays on the network: Westmere cast-iron main phase 3 ($52.0M, $7.9M a year), Cobalt Ridge flood defence ($41.0M, $6.2M), the Harrow Bay 132 kV cable ($68.0M, $9.8M) and the Marlow Vale outlet tower ($33.0M, $4.6M). 57 projects stay funded for $996.0M, removing $383.5M a year. The budget line is now at $1.0B.

Why is the Harrow Bay cable at the line?

It removes a lot of risk, $9.8M a year, but it costs $68.0M: $0.144M of risk removed for each $1M, rank 60 of 214. The value is high because the circuits have leaked oil three times in two years and a release would reach the Harrow estuary — $19.4M of the $54.0M consequence. The business case was in 2024 dollars; $6.3M of escalation is included.

Keep the Westmere main in at $1.0B

It can stay if the two lowest-ranked funded projects are deferred: CP-27-0166 Pinehurst depot fleet charging ($21.0M, $3.7M a year) and CP-27-0177 Ashby Vale 11 kV switchgear ($27.5M, $4.7M a year). The plan costs $999.5M and removes $383.0M a year — $0.5M a year less than the frontier, but it answers Westmere's discoloured-water notice DWN-2026-018. Saved as scenario “$1.0B · Westmere kept”.

Redraft the justification for $1.0B

Redrafting now. The new section names the four projects that drop, the $28.5M a year of risk they carry and the bill difference: +$2.83 a month against +$3.40. Every figure keeps its source.

Every screen

The working solution, as it ships.

13 screens from the working solution, on its sample data. Pick one to see it large.

The efficient frontier214 projects on one curve of risk removed against capital, the budget line at the $1.2B board plan, and what it funds, removes and costs a household.
The optimiser's runEvery step of the run: cases read and put in 2027 dollars, every project valued, must-fund first, seven constraints, the funding line.
Business cases correctedThe 22 cases the reader fixed — escalation, double-counted benefits, missing contingency, overlapping scope — beside their originals.
Projects at the lineThe projects either side of the funding line, with cost, risk removed per $1M, where the risk comes from and the business-case check.
A project's risk caseHarrow Bay's risk built step by step: failures a year × cost of each failure = risk today, then risk removed, with the consequence broken down.
The records behind itThe five failure and condition records used, the business-case checks and the valuer's run, every input cited.
Risk valuesThe value framework that prices every consequence, and the failure curve for each asset class fitted on Lumina's records.
What $1.0B gives upThe budget line moved to $1.0B: four projects drop and the shaded stretch of the curve shows $28.5M a year of risk given up.
A water main below the lineWestmere's cast-iron mains: 61 bursts in five years, the trend, the consequence of each burst and the discoloured-water notice.
Scenarios side by sideThe board plan, the CFO's test and the staff testimony: projects that differ, risk removed by family and the household bill.
The investment justificationRedrafted for $1.0B, with the section on what it gives up, every figure cited in the margin and the regulatory affairs review note.
Sign the recommendationThe recommendation signed with its meaning, the signer and what comes next: the CFO approves the envelope, Regulatory Affairs files.
Your rulesThe envelope by year, must-fund rules, deliverability constraints, approvers and what each agent may do.
Governance

Built for the rate case: cited, signed, on the record.

Every risk value cites its inputsEach project's risk shows the asset register record, the failure records, the class curve and the value framework version behind it, and every figure in the justification links to its source.
Named people decide and signThe Head of Capital Planning signs the recommendation with its meaning recorded; the CFO approves the envelope; Regulatory Affairs files. Agents read, value, rank and draft — they never sign or file.
No signature on a mismatched draftIf the justification is drafted for one envelope and the recommendation is another, the solution says so and will not sign until the draft is redone or the recommendation changed.
Read-only on asset dataThe agents read the asset register, failure records and business cases and never edit them. Lumina's data stays in Lumina's own environment.
Obligations first, rules respectedMust-fund projects — regulatory orders, dam safety directions, signed connection agreements, permit limits — are funded before any ranking. Crew caps, programme floors and dependencies always hold.
Every run and change on the audit trailEach optimiser run, each re-valuation after new failure records, the business-case corrections, comments on a project, the approval of the value framework and each signature are recorded with who and when.
Configuration

Your plan's rules, not ours

The envelope, what is funded first, what the crews can deliver and who signs are settings, chosen by the utility.

SettingDefaultChoose from
Envelope, 2027–2029$1,200.0MSet by year — 2027 $380M · 2028 $410M · 2029 $410M in the sample plan
What happens to the remainder under the lineHeld as contingency (≤ 1 %)Hold the remainder as contingency · Fill it with smaller projects below the line
Must-fund rules, funded before ranking5 of 6 onRegulatory orders and directions · Dam safety inspector's directions · Signed connection agreements · Permit limits with a date · Metering code failures · Board commitments in the last rate case (off)
Water mains crews42 km a yearOn · Off — a deliverability cap the optimiser respects
Cable jointing teams2 major cable projects at a timeOn · Off
Water programmes' share of discretionary spend≥ 30 %On · Off — a programme floor
Dependencies and outage windows4 dependencies · 132 kV outages Apr–Oct onlyOn · Off — e.g. the Northgate second transformer after the interconnection
ApproversHana Okoye · Marcus FeldRecommendation — Head of Capital Planning · Envelope — CFO · Value framework — Dr. Ines Albright · Filing — Owen Hale
Connections

Works with the records a capital plan already runs on

Project business casesPDFs and workbooks — cost by year, scope, assets touched and claimed benefits
Asset register and healthevery asset with its class, age, location, criticality and health score
Failure recordsfaults, bursts, leaks, inspections and near misses, linked to assets
Budget envelopes and the rate modelthe board's envelope by year and the household bill impact
Value frameworkthe dollar value of each consequence — customer-hours, safety, environment, water quality
Regulatory priorities and filingsorders and directions, past capital testimony and data requests
What it changes

The difference, in numbers.

Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.

target
30%
more risk removed for the same capital than ranking by asset age
Ranked by ageOn the frontier
target
100%
of projects in the filing traced to their asset health and failure records
214 of 214 projects
risk value cited
target
Sameday
to show what a budget cut gives up, not a three-week spreadsheet rebuild
Spreadsheet≈ 3 weeks
Budget lineminutes

“demo” = seen in the working solution, on its sample plan data · “target” = the design goal, measured in the pilot · “estimated” = our estimate · “typical” = published figures (EEI: US investor-owned utilities plan about $208B of capital in 2025 and over $1.1T for 2025–2029; Utility Dive via POWER: a record $29B of rate increases requested or secured in H1 2025). Lumina Utilities and the people, places, projects and records named on this page are fictional — characters and sample data in the working solution.

Questions

What capital planners ask us.

What is capital portfolio optimization for utilities?

It is the work of choosing which grid and water projects to fund from a fixed envelope so the plan removes the most risk for the money. Capital Portfolio Optimization does it with five agents that read every business case, value each project's risk from its asset records and draw the efficient frontier; the Head of Capital Planning signs the portfolio.

How does risk-based capital planning work in this software?

Each project's risk is the likelihood of failure of the assets it fixes times the cost of that failure, before and after the work. Must-fund projects are funded first; the rest are ranked by risk removed per dollar within crew caps, programme floors and dependencies, and funded in order until the envelope is used.

Where do the risk numbers come from?

From the utility's own data: health scores in the asset register, failure records, the failure curve for each asset class and a value framework approved by the Director of Asset Management. When an asset has fewer than three records of its own, the class curve is used and the project says so.

Can it show what a budget cut gives up?

Yes. Move the budget line or pick an envelope, and the frontier shows which projects drop and the risk they carried. In the sample plan, moving from $1.2B to $1.0B drops four projects, $194.0M of capital, and leaves $28.5M a year of risk on the network.

Does it help defend the capital plan in a rate case?

The Justification writer drafts the investment justification — how projects were ranked, must-fund obligations, what the plan buys, what a lower envelope gives up and the bill impact — with a source on every figure, and redrafts it for any envelope. Regulatory Affairs edits and files it.

Do people stay in control of the capital plan?

Yes. The agents read, value, rank and draft; they never sign or file and never edit the asset register, failure records or business cases. The Head of Capital Planning signs the recommendation, the CFO approves the envelope and value framework changes need the Director of Asset Management.

Does it work for water as well as electricity?

Yes. Water mains, treatment, wastewater, resilience, growth and electricity network projects are valued on the same scale — customer-hours off supply, properties without water, discoloured-water and boil-water notices, sewage spills, safety and repair — so a burst main and a leaking cable can be compared fairly.

How long does it take to go live?

The Agentic Solution Engine builds and deploys it from your requirements and documents — your business cases, asset register, failure records, value framework and planning rules — and it goes live once every quality gate has passed. We will walk you through it on your own capital plan first.

Supply Chain, Procurement & Finance

More in this division.

All 9 solutions →
Contractor Invoice Validation

Pay for the work done, at the contracted rate.

1-min demoTender Bid Evaluation

Every bid levelled to the same scope, so a hidden exclusion never wins the award.

1-min demoContract Obligation Management

Every penalty owed claimed inside its window, and no renewal missed.

1-min demo

See it on
your capital plan.

We'll run Capital Portfolio Optimization on a sample of your own business cases and asset records.

Book a demoBrowse solutions