How lost time is counted

When work on site is held up, SimOps records it as a Blocker — one record per firm, per stoppage, carrying what happened, when it started, when work restarted, and how many people it stood down. Everything the site later reads as Lost time — the ledger, the weekly report, your firm's own panel — is derived from those records and nothing else. This page explains the counting rules in plain words, so the figures you read (or sign for) hold no surprises. The step-by-step surfaces have their own pages: reading the ledger, your firm's panel, and the walkthroughs.

Three kinds of "lost time" that never mix

Sites lose time in three different senses, and SimOps deliberately keeps them in separate currencies:

• Crew-hours — production lost to stoppages and slowdowns. Declared on blocker records; the subject of this page.
• Person-days — one injured person's time off work, kept on the incapacitation register for statutory reporting.
• Slip — a work line running late against its plan, shown on the work plan's accounting cells.

They appear side by side but are never added together — an hour of lost production and a day of one person's incapacity are different subjects, and any total that blended them would be a fiction. If you remember one design rule, make it this one.

Stopped, or slowed? The record says which

Every record declares its own verb, asked at capture and never guessed:

• Work stopped — the whole gang was effectively stood down.
• Work slowed down — some of the gang kept working.

The verb matters because the numbers must agree with it. If you say work stopped but also say some people kept working, the form stops and asks you to reconcile the two statements — showing what each answer would mean in crew-hours — rather than quietly storing a contradiction. (A record raised from an injury with the answer "don't know" carries no verb at all, and every surface shows not declared rather than inventing one.)

The two questions, and the send-home test

A counted record answers two questions:

1. How big was the gang? Pre-filled from the crews and plant you attach (each attached crew shows how it contributed to the number), and always editable — typing over the figure makes it your firm's declared gang, and the record remembers whether the number was declared or summed from the attached work.
2. How many could you have sent home and still got the same done? This is the send-home test, and it is the number that counts. It deliberately ignores appearances: a gang moved to other work still counts — redeployment doesn't un-lose the time this work lost. Equally, people who were never on this job, or who'd have been away regardless, are not counted.

From there the arithmetic is one line: crew-hours = the stopped window × the people stood down. Nothing else is ever multiplied in — no percentages, no productivity factors, no money. The record shows the crew factor ("3 of 8") and the time factor separately, and never fuses them into a score.

The window: starts when you say, ends when a person says

A record stores when work stopped the moment it's raised. When work restarts is a different kind of fact:

• Only a person can declare a restart — closing the record, or a later correction. No schedule, sweep or timer ever writes one.
• While the record is open it accrues an age ("open for 73h"), and an age is not a loss: a stoppage that hasn't finished counts towards no total, because a figure that grew while nobody was watching would be its own wrong number.
• Close a record while honestly unable to say when work restarted — using the "Nothing to declare — we never restarted" control — and it reads no restart recorded: no window, no hours, stated rather than hidden. The surfaces count how many records are in each of these states instead of quietly showing a smaller total.

What a count rests on

A crew figure has to point at real work. Before a record can carry a count, it must name at least one piece of your firm's work — a crew, a machine, or a scheduled area. That attached work is the record's basis, and it earns its keep three times over:

• it derives the gang default (each attached crew contributes its size, visibly);
• it bounds the window: the form offers your crew's actual site hours as a one-tap suggestion ("Crew on site from 07:00 · use this") — an offer beside the field, never something typed for you;
• it places the loss for the by-location view and the work plan's accounting.

Picking a permit or a Work line expands to the crews placed under it, pre-ticked and strikeable — one tap attaches the whole crew rather than one body at a time. A record that names no work is still always accepted — it just can't carry a crew figure until the work is attached, and it says so plainly.

A stoppage over several days is a chain

Crews go home at night. A stoppage declared Monday 10:00 and closed Wednesday 09:15 did not cost 47 hours of standing crew — so at close, the record offers to split into one record per stretch your crew was actually on site, linked together as a chain. You see the stretches, dated, with three choices: set the real end time, accept the split, or no — we ran straight through. Nothing splits without you choosing, and if the window includes time outside your crew's presence, that share is disclosed on the record rather than silently subtracted. A partial return ("some are back") closes the old record and opens a follow-on with its own freshly-declared count — each record in a chain re-declares the gang the reporter actually has.

Two piles that never mix

The ledger holds records in two tiers of different standing:

• Confirmed — a coordinator has adjudicated the claim: yes, this happened, and (in the same act) answered who was responsible — a firm, or the first-class answer "Nobody — no firm at fault". Only confirmed records carry attribution, and only confirmed records are ever priced.
• Awaiting triage — raised but not yet decided. Its hours are shown as its own total, never added to the confirmed one — silence is not a decision, and a total that mixed vetted and unvetted claims would let a slow triage queue move the site's numbers.

A rejected record is in neither pile: someone decided the loss didn't happen (or wasn't this firm's to claim), with a written reason. It stays readable as history, and the recourse is a fresh raise — the rejection notice itself offers to raise this again, pre-filled.

Three different gestures subtract, and each has its own meaning: correct (the figure is wrong — fix it on the record), reject (the loss didn't happen — reason required), and remove (the record should never have existed — allowed only on a record with no count, no close, no decision, no evidence and nothing pointing at it; anything more established must be rejected with a reason instead, because removing it would take a figure out of the ledger with nothing said).

---

Next: Reading the lost-time ledger · Your firm's lost time · Lost time, scene by scene

Blockers & lost time guide: Overview · Reviewing & triaging · Raising · How lost time is counted · The ledger · Your firm's panel · Slip & nil returns · Standing cost · The register · Scene by scene · Reference