Lifecycle Strategies: Pricing the Work That Defers a Replacement

Every capital forecast you have ever been handed draws the same shape: a line that starts at new and slides down to replacement on schedule. Purchase year, design life, end of life, a cost landing in the year the line hits the floor. It is tidy, it is easy to audit, and nobody actually runs a building or a road that way.

Real programs intervene. Someone recoats the roof before the membrane fails. Someone retubes the boiler instead of buying a new one. Someone crack-seals the collector road the summer before the freeze-thaw gets into it. Each of those jobs costs money and each one moves the replacement date, and in most asset management tools neither fact shows up anywhere except a closed work order.

Lifecycle Strategies, live in AssetLab as of this week, put that work into the model. You describe the interventions once per asset type, and every matching asset projects a curve that bends back up where the work happens. The renewal year moves, the planner re-ranks, and the dashboard tells you what the whole program costs and what it buys.

The assumption baked into every forecast

Straight-line decline is a fine model for a balance sheet. Depreciation has to be simple, defensible, and the same for every asset of a kind, and a straight line is all three. The trouble starts when the same line is asked to do capital planning, because it carries an assumption nobody signed off on: that between the day you buy an asset and the day you replace it, you do nothing.

That assumption has two costs. The first is that your renewal forecast is pessimistic in a way you cannot see. A roof you have been maintaining well will not fail at its design life, but the model says it will, so its replacement sits in a budget year where it does not belong and crowds out something that does. The second cost is worse: the interventions that keep your portfolio alive have no value in the model, so they have no value in the budget conversation. The recoat is an operating expense with no visible return. The replacement it deferred by five years never gets credited to it.

Most teams know this and correct for it by hand. A planner overrides a replacement year because they know the retube happened. A spreadsheet somewhere holds the rule of thumb that sealed roads last two years longer. Those corrections live in people, not in the system, and they walk out the door when the person does.

What a lifecycle strategy is

A lifecycle strategy is the list of interventions you expect to make on an asset between buying it and replacing it, written down once, with enough detail for the model to use. Each intervention is an event with four properties:

• A condition trigger window. The band of condition scores in which the work makes sense. A roof recoat at 70-85, a boiler retube at 40-60, crack sealing at 80-90. Too early and you are spending on an asset that did not need it; too late and the window is missed.
• An effect on condition. Either a number of years added to the asset's life, or a reset to a new condition score. The recoat adds five years; the retube resets the boiler to 100.
• A cost. A fixed figure per event, so the program can be priced.
• A cost source. Where the figure came from: a quote, a contract rate, a past job. This is what makes the number defensible when finance asks.

You set a strategy at the level where the engineering is the same. On the facilities side that is an asset type or a type group: every rooftop unit, every membrane roof, every fire-tube boiler. On the infrastructure side it is a feature class, a material, or a diameter band: every asphalt collector, every 150 mm cast-iron watermain. Set it once and every asset or feature that matches inherits it.

The important word there is inherits. The strategy is a rule about a kind of asset. What each individual asset does with the rule depends on where that asset actually is, which is what the next section is about.

The sawtooth curve

With a strategy in place, an asset's projected condition stops being a straight line and becomes a sawtooth. Condition declines, enters a trigger window, the modelled intervention lifts it back up, and the decline resumes from there. Each tooth is one intervention. The replacement lands where the curve finally reaches your replacement threshold after the last one.

Two details make this more than a prettier chart. First, every asset's curve is anchored at its own assessed condition, not at a theoretical new. Two rooftop units of the same age with the same strategy will project different curves if one was inspected at 82 and the other at 61, and the second one may already be inside its retube window. Second, the same engine runs for both estates. A collector road with a crack-seal-then-resurface strategy produces exactly the same kind of curve as the rooftop unit, so a municipality reads one model across its buildings and its roads.

Where the strategy-aware renewal year goes

The sawtooth produces one number that matters downstream: a renewal year that accounts for the interventions. That number is resolved once and flows into everything that used to read raw design life.

• The lifecycle chart on each asset shows the sawtooth and the strategy-aware renewal year instead of the straight line.
• The replacement planner now ranks by extended life rather than design life, and gains two filters: Due intervention, for assets sitting inside a trigger window right now, and Missed window, for assets that have already declined past one.
• The map's renewal forecast layer reads the strategy-aware year, so a street that will be sealed and resurfaced no longer shows up for renewal in the year its design life happens to expire.
• Corridor dig-once bundling uses the same year when it groups the road, the watermain, and the sewer under it into one project window.

This is also what keeps the Baseline vs Planner forecast honest. The baseline was always "do nothing". Until now the Planner line could only credit replacements; it can now credit the cheaper work that defers them.

Pricing the whole program

A strategy on one asset is a maintenance decision. A strategy across an asset type is a capital program, and it needs to be argued for like one. The dashboard now prices it four ways:

• Year-by-year cash flow for the intervention program, drawn against the run-to-replacement line, so you can see the years where the strategy costs more and the years where it saves the replacement.
• Years of life added across every asset the strategy touches.
• Replacement value deferred: the capital that moves out of the planning horizon, or further into it, because of the work.
• Cost per added year of life, per strategy. Every strategy ranked on the one number that lets a recoat program and a retube program be compared, including what an already-missed window would have cost you.

From trigger window to work order

A model that only forecasts is a report. The point of a trigger window is that it is actionable now: an asset inside one is an asset where the work pays off this year and may not next year. A Due interventions panel lists every asset and feature currently sitting in a window, and turns any of them into a work order or a project in one click.

The Missed window filter is the uncomfortable half of the same idea. An asset that has declined past its recoat band cannot be recoated; the next option is the more expensive one. Seeing that list, next to what the missed window would have cost on the dashboard, is how a program learns which windows it keeps missing and why.

What it will not do

The strategy never writes a condition score. That boundary is deliberate. When the recoat work order is completed, AssetLab offers the modelled post-intervention condition as a pre-filled assessment, and a person confirms it, adjusts it, or rejects it. The model proposes; the inspection decides. Your condition history stays a record of what people observed, not of what a formula expected, which is the only version an auditor or a council will accept.

A strategy is also a statement about a kind of asset, not a promise about one. If a particular boiler has been abused and its retube will not get it to 100, the assessment after the job says so, and that asset's curve re-anchors from the real number.

How to set one up

Start with one asset type where you already know the answer. Most facilities teams have a roofing or HVAC program they run from memory; most public works departments have a pavement preservation sequence. Write that down first.

• Pick the type or class. An asset type or type group on the facilities side; a feature class, material, or diameter band on the infrastructure side.
• Add the interventions in order, each with its trigger window, its effect (years added or condition reset), its cost, and where the cost came from.
• Check that the matching assets have an assessed condition. The curve anchors at that score, so an asset without one has nothing to anchor to.
• Open the plannerand filter on Due intervention. That list is your first quarter's work.
• Read the dashboardfor the program's cash flow and cost per added year. That is the slide for the budget meeting.

If you manage strategies at scale or want them in your own systems, lifecycle events are available as full resources in the REST API and MCP server, so a consultant's pavement model or an engineering standard can be loaded rather than re-typed.