Commit in stages, with a rule the committee can actually delegate.
The original model favours £650k of phasing now, with expansion only if signal G appears at the end of Y2. The future demand state remains unknown at that gate. This page compares executable policies and the cash authority needed to follow them.
Original decision record
This fixed record is retained when sensitivities change. The controls create temporary calculations in this browser only. Reload or Reset restores the original assumptions; nothing is saved, submitted, funded or approved.
Test the committee’s assumptions
Choose one of the specified savings, joint forecast mix or liquidity cases. All three policies and their gate choices are recalculated immediately. F7
The investment committee must authorise the initial structure, signal-to-action delegation, cash ceiling and grant conditions. Leena recommends; these controls do not exercise that authority. F8
Compare the three investment policies
Each P/D rule selects the better conditional future value at each signal, subject to the whole-branch cash ceiling. Full conversion is a fixed commitment. All amounts below are £000; NPV includes all capital, operating cash and earned grants. F1 F6
| Initial policy | Observable gate rule | Expected NPV | Worst branch NPV | Peak draw | Expected grant | Funding status |
|---|
Expected grant is probability-weighted cash received, not money available at t=0. Full: £150k at end Y2 under G/R only. Phase: £100k at end Y3 only for a G/R expansion commissioning in Y3. Wait: no grant. F5
The gate is an information boundary
Why each rule chooses its action
The two continuation values are measured at end Y2, immediately before gate capital and after Y2 operations. They exclude already-incurred cash, include gate capital, and discount Y3–Y6 operating cash and grants back to the gate. A positive investment increment alone is insufficient: funding must work in both remaining states. F1 F6
| Structure / signal | Unresolved states Conditional weights | Investment quote | Decline value | Invest value | Invest less decline | Peak if invest | Implemented action |
|---|
Exact ties decline investment. Capital and commissioning are specific to each structure; no option borrows another option’s price. Changing g preserves the within-signal weights, so g can change the initial policy without changing the conditional gate economics. F2 F7
Follow the money through all six futures
Draw is the positive cash shortfall, max(0, −cumulative net cash). Overlapping traces are deliberate where the cash positions match. The Y2 receipts and capital points are shown separately; annual trails below retain any cash surplus. F6
| State | Probability | Gate action | NPV | Peak draw | Grant cash | Final cumulative cash |
|---|
The eight requested sensitivity cases
These are separate assumptions, with no probabilities assigned across settings. Every row re-optimises the gate rule under its own cash ceiling. Use Inspect to reproduce the row above. F7
| Assumptions | Full now | Phase now | Wait | Preferred policy | Explore |
|---|
What reverses the choice: at the original 35% G mix, 25% lower savings makes wait/no investment preferable; at 90% G, phasing still pays at that savings level, while full conversion leads at u=1.00 or 1.15. Tightening cash authority to £1,050k excludes full conversion because the A branches draw £1,070k before commissioning. It also blocks wait/build-on-G (£1,100k draw). At u=1.00, phasing is consequently preferred at both mixes. F3 F6 F7
The base G expansion decision is fragile: its £13.6k conditional gate value assumes the supplied joint demand forecast, commissioning and grant timing. Actual future demand remains uncertain, including a negative-NPV G2 outcome. A materially changed quote, grid slot or credible new evidence requires a new appraisal and appropriate authority, not an invented distinction between G1 and G2. F3 F8
Model workings and assumptions
Estimating basis and terms: six-year incremental real cash flows against continuing diesel; tax and financing excluded; 8% discount rate; year-end operations; no terminal value because the lease ends after Y6. Capital already includes net end-of-lease obligations and excludes common vehicle replacement costs. No value is given to unserved demand. F1
Forecasts, not observed outcomes: the demand paths, joint state probabilities, connection timings and price relationships come from the supplied planning model. The gate observations have not happened. Authorised estimating inputs do not amount to investment approval. F2 F3 F8
F standing cost is £45k in every year, including before commissioning. P costs £25k until capacity 60 operates, then £45k. D costs nothing before gate investment; after investing, any whole waiting year costs £25k and operating years cost £45k. A gate build starts in Y4; G/R starts in Y3. F4
Current joint forecast
P(G1)=4g/7; P(G2)=3g/7; P(A1)=4(1−g)/13; P(A2)=P(R1)=P(R2)=3(1−g)/13. This preserves the supplied dependencies; it does not recombine demand, grid and price independently. F3 F7
| State | Gate signal | Probability | Demand Y1–Y6 | Full 60 starts | Gate 60 starts |
|---|
Supplied fictional evidence
The records below are the source text for this model. Scenario controls leave them unchanged. Calculated results elsewhere are derived findings, not additional source facts.
F1 — Appraisal instruction, 1 November 2027; finance director.
t=0 is 1 January 2028; Y1–Y6 are 2028–2033. All money is £000 in constant prices, incremental to continuing diesel, excluding tax/financing. Discount rate 8%; annual operating receipts/costs occur at year-end. Initial capital is t=0. Gate capital is paid at end Y2 after that year's operating cash and any grant receipt. No terminal value: the depot lease ends after Y6 and quoted capital includes the net end-of-lease obligations. Vehicle replacement costs common to diesel are already netted out. No value is assigned to unserved demand.
F2 — Engineering options, 2 November; authorised estimating basis.
| Initial choice | t=0 payment | Before the Y2 gate | Permitted gate choice |
|---|---|---|---|
| Full F | 980 | Install for capacity 60; commissioning depends on the state below | No redesign/cancellation choice in this comparison |
| Phase P | 650 | Capacity 30 from start Y1 | Keep capacity 30 for remaining lease, or pay the signal's expansion quote for capacity 60 |
| Wait D | 0 | Capacity zero; continued diesel baseline | Do nothing for remaining lease, or pay the signal's full-build quote for capacity 60 |
F has zero electric capacity before its stated 60-capacity commissioning year. P retains 30 until expansion commissions; D has zero until its gate investment commissions. No other interim conversion is available. The gate is the sole later investment opportunity; a declined gate investment cannot be made in Y3/Y4 instead. Engineering has confirmed the quoted construction/commissioning patterns. These are alternative project structures, not interchangeable purchase prices.
F3 — Joint commercial/grid forecast, 4 November; fleet and connection managers.
Six mutually exclusive and exhaustive planning states. Demand is annual average eligible vehicles. One state's probability covers its demand path, grid and price jointly; do not recombine marginal scenarios.
| State | Probability | Gate signal | Demand Y1,Y2,Y3,Y4,Y5,Y6 | First 60-capacity year if F | First 60-capacity year if gate investment | P expansion quote | D full-build quote |
|---|---|---|---|---|---|---|---|
| G1 | .20 | G | 20,45,55,60,60,60 | 2 | 3 | 500 | 1,100 |
| G2 | .15 | G | 20,45,30,25,25,25 | 2 | 3 | 500 | 1,100 |
| A1 | .20 | A | 20,25,50,55,60,60 | 3 | 4 | 800 | 1,400 |
| A2 | .15 | A | 20,25,20,20,20,20 | 3 | 4 | 800 | 1,400 |
| R1 | .15 | R | 10,15,35,40,45,50 | 2 | 3 | 600 | 1,200 |
| R2 | .15 | R | 10,15,15,15,15,15 | 2 | 3 | 600 | 1,200 |
Demand paths are estimates, not commitments. Gate signals are observed only at the end of Y2: actual Y1/Y2 demand, the firm connection slot and investment quote distinguish G/A/R. They do **not** distinguish G1 from G2, A1 from A2 or R1 from R2. No forecast signal about that within-group future becomes available before the gate decision. The same information is available under F/P/D; ordinary fleet forecasting has no extra project cost.
F4 — Operating model, 5 November; operations finance.
Vehicles served in a year = min(demand, commissioned capacity). Each served vehicle saves £9k/year before standing cost, multiplied by a common savings factor u. Use u=1 initially. This factor changes savings, not demand or capacity. F standing cost is £45k every year, including years before commissioning. P costs £25k/year until capacity 60 operates, then £45k/year; no expansion leaves £25k throughout. D costs zero before investment; after gate investment it costs £25k in any whole waiting year and £45k from the first operating year. Commissioning occurs at the start of the indicated year. No other operating cost is omitted.
F5 — Grant terms, 5 November; grant manager.
F receives £150k at end Y2 only if capacity 60 is operational from start Y2; delayed F gets no grant. P receives £100k at end Y3 only if its stage-one commitment was made at t=0 and its expansion produces capacity 60 from start Y3. A delayed expansion gets no grant. D gets no grant. No grant is received before its condition is met. Grants do not alter physical cost or commissioning dates.
F6 — Liquidity rule, 6 November; treasury, binding.
£1,200k is the maximum cumulative net project cash draw approved at any modelled payment point, not a capital-cost or NPV ceiling. Operating savings and grants remain available to the project; calculate draw from all cumulative undiscounted cash flows, including t=0. If a gate action would breach the ceiling in either still-possible state, that action is not fundable on existing authority. Do not assume borrowing, net positive expected cash instead of branch cash, or use discounted cash for this test. No cash flows occur between the stated annual payment points. Treasury can consider a new request, but none is approved.
F7 — Committee sensitivity request, 8 November.
Test u=.75, 1.00 and 1.15. Also test total probability g of signal G at .35 and .90: within G preserve G1:G2=4:3; allocate remaining probability to A:R=7:6, preserving A1:A2=4:3 and R1:R2=1:1. Base g=.35 reproduces F3. Separately test a £1,050k liquidity ceiling. Recompute gate decisions under changed assumptions; do not hold a policy fixed and describe it as re-optimised. Required comparison: all six combinations of u={.75,1.00,1.15} and g={.35,.90} at the base liquidity ceiling, then u=1.00 with g=.35 and .90 at the £1,050k ceiling. No probabilities are assigned to these sensitivity settings.
F8 — Authority and decision standard, 10 November; committee chair.
Leena recommends; the investment committee authorises the initial choice, gate delegation, cash ceiling and grant conditions. Prefer the highest expected NPV among fundable policies on the stated monetary model, while disclosing branch losses and the estimates that could reverse the decision. A cautious alternative may be recommended with its quantified opportunity cost and rationale. A later gate may be delegated only as an observable signal-to-action rule. Preserving the option to make a different recommendation after genuinely new evidence is not permission to use future state information now.
Read the complete supplied case · BPB2-02
# BPB2-02 — Freight depot electrification — make the investment case
Suite BPB2 v0.3-pilot; Foray 247, third essai.
### Scenario
Northmere Freight must decide how to electrify a leased depot whose operation ends in six years. Committing now can secure a grant, while a smaller first phase buys early savings and a later decision. Future demand, grid timing and quoted expansion cost are related, and the later gate reveals only part of that future. Commercial director Leena Orr needs an investment policy the finance committee can fund and actually follow, not the value of a clairvoyant choice among completed scenarios.
### Exact task prompt
Produce one usable in-browser answer page for the named practitioner, using the supplied fictional evidence. Keep observed facts, forecasts, assumptions and proposed decisions distinguishable, and cite source IDs beside material claims. No external research or real-world certification is required.
Prepare Leena's six-year incremental investment case and recommend an implementable policy: full conversion now, initial phasing with a later expansion rule, or waiting with a later investment rule. Compare expected NPV, branch outcomes, grants and peak funding draw under the supplied information structure. Show what is decided now, what can be decided at the gate and which future uncertainty remains unresolved. Test the specified savings, scenario-mix and liquidity sensitivities, including changes in preferred policy. Make intermediate cash flows inspectable and distinguish economic desirability from funding feasibility. A single base-case spreadsheet or a different decision for two states that look identical at the gate is insufficient.
### Supplied records
**F1 — Appraisal instruction, 1 November 2027; finance director.** t=0 is 1 January 2028; Y1–Y6 are 2028–2033. All money is £000 in constant prices, incremental to continuing diesel, excluding tax/financing. Discount rate 8%; annual operating receipts/costs occur at year-end. Initial capital is t=0. Gate capital is paid at end Y2 after that year's operating cash and any grant receipt. No terminal value: the depot lease ends after Y6 and quoted capital includes the net end-of-lease obligations. Vehicle replacement costs common to diesel are already netted out. No value is assigned to unserved demand.
**F2 — Engineering options, 2 November; authorised estimating basis.**
| Initial choice | t=0 payment | Before the Y2 gate | Permitted gate choice |
|---|---:|---|---|
| Full F | 980 | Install for capacity 60; commissioning depends on the state below | No redesign/cancellation choice in this comparison |
| Phase P | 650 | Capacity 30 from start Y1 | Keep capacity 30 for remaining lease, or pay the signal's expansion quote for capacity 60 |
| Wait D | 0 | Capacity zero; continued diesel baseline | Do nothing for remaining lease, or pay the signal's full-build quote for capacity 60 |
F has zero electric capacity before its stated 60-capacity commissioning year. P retains 30 until expansion commissions; D has zero until its gate investment commissions. No other interim conversion is available. The gate is the sole later investment opportunity; a declined gate investment cannot be made in Y3/Y4 instead. Engineering has confirmed the quoted construction/commissioning patterns. These are alternative project structures, not interchangeable purchase prices.
**F3 — Joint commercial/grid forecast, 4 November; fleet and connection managers.** Six mutually exclusive and exhaustive planning states. Demand is annual average eligible vehicles. One state's probability covers its demand path, grid and price jointly; do not recombine marginal scenarios.
| State | Probability | Gate signal | Demand Y1,Y2,Y3,Y4,Y5,Y6 | First 60-capacity year if F | First 60-capacity year if gate investment | P expansion quote | D full-build quote |
|---|---:|---|---|---:|---:|---:|---:|
| G1 | .20 | G | 20,45,55,60,60,60 | 2 | 3 | 500 | 1,100 |
| G2 | .15 | G | 20,45,30,25,25,25 | 2 | 3 | 500 | 1,100 |
| A1 | .20 | A | 20,25,50,55,60,60 | 3 | 4 | 800 | 1,400 |
| A2 | .15 | A | 20,25,20,20,20,20 | 3 | 4 | 800 | 1,400 |
| R1 | .15 | R | 10,15,35,40,45,50 | 2 | 3 | 600 | 1,200 |
| R2 | .15 | R | 10,15,15,15,15,15 | 2 | 3 | 600 | 1,200 |
Demand paths are estimates, not commitments. Gate signals are observed only at the end of Y2: actual Y1/Y2 demand, the firm connection slot and investment quote distinguish G/A/R. They do **not** distinguish G1 from G2, A1 from A2 or R1 from R2. No forecast signal about that within-group future becomes available before the gate decision. The same information is available under F/P/D; ordinary fleet forecasting has no extra project cost.
**F4 — Operating model, 5 November; operations finance.** Vehicles served in a year = min(demand, commissioned capacity). Each served vehicle saves £9k/year before standing cost, multiplied by a common savings factor u. Use u=1 initially. This factor changes savings, not demand or capacity. F standing cost is £45k every year, including years before commissioning. P costs £25k/year until capacity 60 operates, then £45k/year; no expansion leaves £25k throughout. D costs zero before investment; after gate investment it costs £25k in any whole waiting year and £45k from the first operating year. Commissioning occurs at the start of the indicated year. No other operating cost is omitted.
**F5 — Grant terms, 5 November; grant manager.** F receives £150k at end Y2 only if capacity 60 is operational from start Y2; delayed F gets no grant. P receives £100k at end Y3 only if its stage-one commitment was made at t=0 and its expansion produces capacity 60 from start Y3. A delayed expansion gets no grant. D gets no grant. No grant is received before its condition is met. Grants do not alter physical cost or commissioning dates.
**F6 — Liquidity rule, 6 November; treasury, binding.** £1,200k is the maximum cumulative net project cash draw approved at any modelled payment point, not a capital-cost or NPV ceiling. Operating savings and grants remain available to the project; calculate draw from all cumulative undiscounted cash flows, including t=0. If a gate action would breach the ceiling in either still-possible state, that action is not fundable on existing authority. Do not assume borrowing, net positive expected cash instead of branch cash, or use discounted cash for this test. No cash flows occur between the stated annual payment points. Treasury can consider a new request, but none is approved.
**F7 — Committee sensitivity request, 8 November.** Test u=.75, 1.00 and 1.15. Also test total probability g of signal G at .35 and .90: within G preserve G1:G2=4:3; allocate remaining probability to A:R=7:6, preserving A1:A2=4:3 and R1:R2=1:1. Base g=.35 reproduces F3. Separately test a £1,050k liquidity ceiling. Recompute gate decisions under changed assumptions; do not hold a policy fixed and describe it as re-optimised. Required comparison: all six combinations of u={.75,1.00,1.15} and g={.35,.90} at the base liquidity ceiling, then u=1.00 with g=.35 and .90 at the £1,050k ceiling. No probabilities are assigned to these sensitivity settings.
**F8 — Authority and decision standard, 10 November; committee chair.** Leena recommends; the investment committee authorises the initial choice, gate delegation, cash ceiling and grant conditions. Prefer the highest expected NPV among fundable policies on the stated monetary model, while disclosing branch losses and the estimates that could reverse the decision. A cautious alternative may be recommended with its quantified opportunity cost and rationale. A later gate may be delegated only as an observable signal-to-action rule. Preserving the option to make a different recommendation after genuinely new evidence is not permission to use future state information now.
### Browser deliverable and completion conditions
An investment decision page exposing each implemented signal rule, six branch cash-flow trails, expected NPV, peak draw and grant receipts. Changing u, g or liquidity must recompute policy feasibility and the recommended policy comparison, including gate choices. Allow the reader to inspect the two future states under a single gate signal and see the same action applied. Preserve the original decision record and identify hypothetical changes. Include a clear return to the gallery; exports cannot substitute for working calculations.
The page must work in the supplied browser/runtime; a download button or file preview alone is insufficient. Include a clear “Return to gallery” link using the shared shell route, or ../index.html if no route is supplied. Required controls must change the dependent results, not just their appearance. Original evidence and the initial dated position must remain recoverable; hypothetical changes do not create approvals or alter source records. Optional exports are secondary.