Design versions and assumptions
The key idea
A result belongs to the inputs that produced it. Changing a draft does not change an old result, and an entered value does not become verified evidence simply because it was saved.
Three records, three roles
The draft is what you are editing. A saved version is a snapshot that can be calculated again. A run records the calculation against a particular version and its assumptions.
A report should identify the run it uses. Otherwise a plausible chart can be paired with a different equipment size, demand year or price assumption.
The common mistake is to edit a field, look at an existing result and assume the result now includes that edit.
Make a result old on purpose
Start by calculating the saved version. Then change the solar capacity in the draft. The result keeps its original amount.
Save the new version. The old run still exists, but it no longer belongs to the selected saved version. Calculate again to obtain a current result.
The Adopt 60 kW candidate action illustrates promotion into a new saved version. It is a declared teaching candidate, not an optimizer recommendation.
01 · Draft
40 kW
Matches saved
02 · Version 1
40 kW
Saved input snapshot
03 · Result
No result
Calculate the saved version
No calculation is recorded.
Read the values and assumptions
| Field | Saved input | Result receipt |
|---|---|---|
| Solar capacity | 40 kW | No result |
| Unit cost | 800 USD/kW | No result |
| Origin | assumed | No result |
| Evidence status | No supplier source attached | No result |
| Derived capital | Calculated from saved values | No result |
Model note · Deterministic teaching records, not a Phasor project. The calculation is only PV capital = kW × USD/kW, not an annual simulation or project valuation. Initial rate $800/kW is an unsourced teaching assumption. Marking it user-entered records authorship, not verification. Saving retains a new snapshot; calculating reads that saved snapshot, never the unsaved draft.
The two status questions are different: does the result match the saved version, and does it describe the unsaved draft? A result can be current for the saved version while a new draft waits.
In Phasor, some actions save and run in one operation. This explorer separates the operations so the distinction is visible.
Read an assumption receipt
| Origin | What it tells you | What it does not prove |
|---|---|---|
| Default or assumed | The software supplied a starting value. | That it applies to this project or has a complete source. |
| User-entered | A person selected or entered the value. | That the value was measured, quoted or independently checked. |
| Derived | A calculation produced it from other inputs. | That those inputs were correct. |
The initial 800 USD/kW in the explorer has no supplier source. Changing its label to user-entered records a choice. It does not invent a quotation, source date or region.
Missing evidence remains named in the result receipt. A review should distinguish “the arithmetic can run” from “the assumptions are adequate for this decision.”
The math, if you want itOptional — the page reads completely without it
The teaching calculation is deliberately trivial, so the record-keeping is the whole lesson:
the only formula in the explorer
capital = Psolar · cunit
version 1
capital = 40 kW × 800 USD/kW = 32,000 USD
Editing the draft to 50 kW leaves that recorded result at 32,000 USD. After saving and calculating the new version, it becomes 50 × 800 = 40,000 USD, explained entirely by the input change; no economic return or annual operating result has been calculated.
If the provenance changes while the number stays 800, the product is unchanged. The evidence record still changes and travels with the new version.
Promote with a traceable comparison
When choosing a sizing alternative, keep its source run and candidate identity. Inspect the limits and assumptions it satisfied, save the adopted design, and simulate the chosen version.
For a coupled network sizing proposal, model revisions also matter. If the underlying model changes before application, recompute the proposal. Re-entering stale proposed numbers manually removes the protection provided by that revision check.
Before you export
Confirm the design, version, run, demand basis, weather provenance, cost currency and time horizon. Read unresolved assumptions with the headline result. A screenshot that omits identity and caveats can hide an important difference even when every displayed number is genuine.
Financial scenarios have their own saved-document behaviour. They are not reversed by the ordinary model Undo command. Keep a separate scenario when you need to preserve a financial decision.
See it in Phasor
Use the version bar and selected run identity, then read Results → Assumptions. Agent-marked fields can still be unsaved drafts. Confirm the save or apply action before expecting a new report to include them.
Continue
Use the AI assistant guide for staged edits and approvals, or the Financial model guide for scenario persistence.