docs: describe solution T through the cash flow, S53 and S54

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
This commit is contained in:
Kjell Tore Guttormsen 2026-11-13 00:37:50 +01:00
commit e0db6ab4b7
Signed by: ktg
SSH key fingerprint: SHA256:JakMjO6FTBBzN0Bhfj9saOoEjaFxlSdYuZQQpM/lF9Q
3 changed files with 37 additions and 13 deletions

View file

@ -84,3 +84,15 @@ All notable changes to this project will be documented in this file.
made-up sets for S51–S54 without a difference. Six clarifications (70–75) in chapter 16 of the specification
say how chapter 17 is read where the text was silent. Solution T is not taken through the cash flow or the
verdicts yet, and no result for the hypothesis has been computed.
- Solution T through the power system and the cash flow (`kontantstrom.py`): everything per module, renewable
modules with their own portfolio as in C, the nuclear modules of T-K on shared units as in B, and the fixed
cost of a new nuclear country counted to the first unit. The loss of demand of chapter 17.3 enters the
revenue and the disconnection arm of S47 (`regn_program(..., bortfall=...)`); without a loss every figure is
unchanged. Emissions "with deduction" (S53) and cooling water (S54) are put together over a programme for
all seven solutions (`miljo.py`). A new independent reader coded the same from the specification alone:
1 778 made-up paths through both codes without a difference, and the two sensitivities that are switches in
both. Six clarifications (76–81) in chapter 16 settle what the text left open
(`modell/fagfellevurdering-utkast-2-og-3.md`, part 11). The levelised cost, ownership, scale and verdicts
refuse solution T and a loss of demand until chapter 17 is taken through them. No result has been computed.
- Present values are now exactly rounded sums. As dot products the same computation did not give the same last
digit every time, for every solution. A test runs the same computation six times and requires equal numbers.

View file

@ -19,8 +19,9 @@ carousel, Python code, and a Claude plugin (own repo, later).
Every model parameter must trace to an entry here.
- `FORHANDSREGISTRERING.md`: the approved direction and pre-registered criteria (M1–M7). Committed.
- `modell/spesifikasjon.md`: the model specification (equations S1–S54; of chapter 17,
the addition to the direction, the build-out of solution T and the equations S48–S54 are coded, the rest
is not). `modell/parametere.json`: the frozen
the addition to the direction, solution T is coded through the build-out, the power system and the cash
flow, and S48–S54 are coded; T through the levelised cost, ownership, scale and verdicts is not).
`modell/parametere.json`: the frozen
parameters (value, range, distribution, source entries); its checksum is recorded in the pre-registration
(version 1.1 after the one deviation of 2026-11-12, which added the quantities of chapter 17.8).
The table in chapter 14 of the specification is generated from that file.
@ -45,12 +46,18 @@ carousel, Python code, and a Claude plugin (own repo, later).
ownership forms (residual value, levies, private owners, the rate the state's cash flow is valued with, the
cost of capital before tax), approved by the operator 2026-11-12; 67 lists what remains. Clarification 68
holds the sources' figures for M5 (i), mirrored in `miljo.KILDEVERDIER_M5`. Clarifications 70 to 75 hold the
readings of chapter 17 made when its build-out was coded, settled against an independent reader's code.
Solution T (`Losning.T_F`, `Losning.T_K`) only has its build-out: `lag_plan` returns a `Trinnvisplan`, and
`regn_program` refuses T until it is taken through the power system, cash flow, ownership and verdicts.
S51 to S54 are pure functions in `ligninger.py`; the factor of S51 is not in the revenue yet, and S53 and
S54 are not put together over a programme. The run over the draws, the sensitivities and the tools of
chapter 12 are not coded yet.
readings of chapter 17 made when its build-out was coded, and 76 to 81 those made when T was taken through
the power system and the cash flow, both settled against an independent reader's code.
Solution T (`Losning.T_F`, `Losning.T_K`) goes through `lag_plan` (a `Trinnvisplan`) and `regn_program`
with a module in the place of a facility (`Programresultat.P_enhet`): renewable modules have their own
portfolio as in C, the nuclear modules of T-K share the programme's units as in B
(`paa_felles_kjernekraft`), and S41 counts to the first unit. The factor of S51 enters the revenue and the
disconnection arm of S47 through `regn_program(..., bortfall=...)` (`Programresultat.mu_d`; exactly one
without a loss). `miljo.py` puts S53 (`unngaatt_utslipp`, `utslipp_med_fradrag`) and S54 (`kjolevann`)
together over a programme for all solutions. `nivaaisert.py`, `eierskap.py` and `skala.py` refuse a result
for T or with a loss of demand (`ks.krev_uten_tillegget`) until part 5T-4c takes chapter 17 through them.
Present values are exactly rounded sums (`math.fsum`), never numpy dot products: those are not repeatable
to the last digit. The run over the draws, the sensitivities and the tools of chapter 12 are not coded yet.
- `data/trekninger.npy` and `data/trekninger.json`: the 10 000 draws of the 91 drawn quantities, in the parameter
file's units. `data/aarshendelser.npy` and `.json`: one uniform number per draw and year 2028–2100 for the
export rule (own seed). Committed; every later computation reads them.
@ -58,7 +65,8 @@ carousel, Python code, and a Claude plugin (own repo, later).
NOU 2026: 4, Epoch and NVE with the sources' own assumptions). Build-out, power system and cash flow are tested
on the made-up set of values in `tests/conftest.py`, and so are the levelised cost, ownership, scale and
environment. The verdicts are tested on invented outcomes per draw. `test_trinnvis.py` holds chapter 17:
hand-computed build-outs for T-F and T-K, and S48 to S54.
hand-computed build-outs for T-F and T-K, and S48 to S54. `test_trinnvis_kontantstrom.py` holds T through
the power system and the cash flow, the loss of demand, and S53 and S54 over a programme.
- `requirements.txt` (numpy, scipy) and `requirements-dev.txt` (pytest, ruff, mypy) pin exact versions;
`pyproject.toml` only configures the tools (ruff `ALL`, mypy strict).

View file

@ -15,7 +15,8 @@ skal si når satsingen lønner seg, og når den ikke gjør det.
- `modell/spesifikasjon.md`: modellen: ligninger, størrelser og parametere med kilde. Kapittel 16 har
avklaringene som ble gjort etter at en uavhengig leser kodet ligningene. Kapittel 17 har tillegget til
retningen: trinnvis oppbygging, AI-bruk på brukernes egne maskiner, utslipp «med fradrag» og kjølevann. Av
det er utbyggingen i moduler og ligningene S48–S54 kodet; resten er ikke kodet ennå.
det er utbyggingen i moduler, kraftsystemet, kontantstrømmen og ligningene S48–S54 kodet; resten er ikke
kodet ennå.
- `modell/parametere.json`: parameterne, frosset før noe er regnet: verdi, spenn, fordeling og poster i
kildebanken. Fila er endret én gang etter frysen, 12. november 2026, da størrelsene for tillegget til retningen
kom inn, også det før noe er regnet. Sjekksummen og hva som ble endret, står i forhåndsregistreringen.
@ -24,9 +25,12 @@ skal si når satsingen lønner seg, og når den ikke gjør det.
- `src/tokenfabrikker/`: koden. Én funksjon per ligning, med samme nummer som i spesifikasjonen: kraftverket,
datasenteret og inntekten (kapittel 1–3), utbyggingen (kapittel 4), kraftsystemet (kapittel 5), kontantstrømmen
med netto nåverdi for Norge (kapittel 6), bedriftsøkonomien, eierformene, skala, miljø og dommene (kapittel
7–11), og innlesing av parameterne og trekningene. For den trinnvise løsningen er bare utbyggingen kodet
(kapittel 17.1): når hver modul starter, og hvilken kraft den går på. Kjøringen over trekningene,
følsomhetene og verktøyene er ikke kodet ennå, og ingen resultater er regnet.
7–11), og innlesing av parameterne og trekningene. Den trinnvise løsningen (kapittel 17) er kodet til og med
kontantstrømmen og netto nåverdi for Norge: når hver modul starter, hvilken kraft den går på, og hva
programmet koster og tjener år for år. Utslipp «med fradrag» og kjølevann regnes for alle løsningene, og
lavere etterspørsel etter regnekraft kan legges inn i inntekten. Kraftkostnaden per årgang, eierformene,
skala og dommene er ikke ført gjennom for den trinnvise løsningen ennå. Kjøringen over trekningene,
følsomhetene og verktøyene er ikke kodet, og ingen resultater er regnet.
- `data/trekninger.npy`: de 10 000 trekningene av parameterne, laget med fast frø. `data/trekninger.json` sier hva
kolonnene er. `data/aarshendelser.npy` er en egen strøm av tilfeldige tall for eksportregelen, ett tall per
trekning og år.