A replacement decision asks: should we keep using the existing asset (the defender), or replace it with a new asset (the challenger)? Engineering-economic analysis answers this by computing the EAC of each option and picking the lower one — but the framing has several subtleties unique to replacement analysis.
Reasons to replace or retire an asset:
-
Replacement — substituting a new asset for the old one — happens when the defender’s performance has degraded (rising maintenance costs, lower output, more downtime) or when a better technology has appeared (lower operating cost, higher capacity, new capabilities).
-
Retirement — removing the asset entirely without replacement — happens when the service the asset provided is no longer needed.
The decision is recurring: defender vs. challenger is re-asked each year as long as the defender remains in service. New challengers may appear, the defender keeps aging, and what was the right call last year may not be the right call this year.
Economic life vs. physical life vs. service life
- Physical life is how long the asset can run before it physically falls apart. Can be decades.
- Service life is how long the asset is in service — could be less than physical life if it’s retired early.
- Economic life is the optimal life from a cost-minimisation perspective — the age at which the EAC of keeping the asset is at its minimum.
Economic life is usually shorter than physical life. As an asset ages, capital cost amortised per year (which falls — you’ve owned it longer) trades off against operating-and-maintenance cost per year (which rises — the asset is wearing out). The optimal hold is where the EAC bottoms out. Holding the asset longer than its economic life is overall more expensive even though the capital is “free” by then.
Key principle: ignore sunk costs
The defender’s original purchase price is a sunk cost — irrelevant. What matters for the keep-or-replace decision is the defender’s current market value (the opportunity cost of not selling it now) and its future cash flows (O&M costs, eventual salvage).
In replacement analysis, the defender’s “first cost” used in EAC calculations is the current market value, not the historical purchase price. The market value is what you’d get for selling it today — that’s the value you’re giving up by keeping it.
Replacement scenarios
Three scenarios show up:
Scenario 1: Identical defender and challenger (asset need is indefinite, life cycle repeats).
The challenger today is identical to the defender, and any future replacement will also be identical. Stable technology, stable prices, stable interest rates. The decision rule: replace when EAC_capital is minimised, i.e., when the defender reaches its economic life. The exact replacement time follows directly from the EAC vs. age curve.
Scenario 2: Different defender and challenger (same challenger continues indefinitely).
You have an old defender, a new challenger, and assume any future replacement will be a new copy of the challenger. The decision rule:
- Find the economic life of the challenger and its EAC_c at that life.
- Compute the defender’s EAC_d for each remaining year of its useful life.
- If EAC_d > EAC_c at any point, replace now.
- Otherwise, monitor: as the defender ages, EAC_d will eventually rise above EAC_c at year . Replace at year — the last year before the defender is more expensive.
Scenario 3: Different defender, different future challengers.
The future challengers are expected to be technologically different (cheaper, more capable, less polluting). This is the most realistic case for tech-heavy assets — but also the most analytically complex. It requires forecasting future challenger economics, which is largely guesswork. Not covered in Engineering Economics.
One-year principle
For assets where capital costs are small relative to O&M costs, and O&M costs increase steadily with age, the economic life of the defender simplifies to one year: the marginal year-of-keeping is the right comparison. EAC_total reduces to EAC_O&M evaluated at .
This is a handy shortcut for old, paid-off assets whose decision really is “one more year, or replace?” The cleaner the simplification, the more useful it is — and for assets whose capital was sunk long ago, the simplification is essentially exact.
For the cost-component breakdown see Equivalent annual cost. For the time-horizon concept see Economic life. For the underlying decision framework see Comparison of alternatives and Sunk cost.