Get Smart With Critical Spares

Critical-spare decisions reflect risk management under uncertainty. Comparing the cost of failure with and without a spare can ease your mind.

By Drew D. Troyer, CRE, Contributing Editor

One of the toughest issues in managing reliability is that of critical—or insurance—spares. Unlike consumables and ordinary repair items, where bin-level and economic-order-quality (EOQ) processes are effective, critical spares present a challenge in that the assets they support aren’t expected to fail. This means critical-spare decisions are basically risk management under uncertainty. My approach involves a straight-up economic comparison of the cost of failure with and without the spare, which includes the cost of inventory management, cost of money, and time value of money. Let’s explore this further.

Assume a machine has a failure cost of $80,000 per event, including parts, labor, downtime, and risk-based costs, factoring in lead-time on a targeted part if it’s not spared. Also assume that the same failure cost drops to $40,000 if the part is spared, creating a $40,000 differential between the two scenarios. Let’s also assume that the failure rate is 0.05 failures/yr., yielding a mean-time-between-failure (MTBF) of 20 years. Additionally assume that the cost to purchase the spare is $30,000. Sounds relatively straightforward, right? A $30,000 investment yields $2,000 in mitigated risk per year, right? Not so fast. There are a number of other considerations:

• How good do you feel about your estimates of failure cost and MTBF? For a reality check, I usually use sensitivity analysis to evaluate best-, likely, and worst-case scenarios.

• How many assets will one critical spare protect? The investment looks very different if a single critical spare serves, say, five systems. When evaluating scenarios where one spare serves multiple assets, consider the conditional probability that two assets will fail and demand the same spare at the same time.

• What’s the inventory’s carrying cost? In most plants, it’s about 22% of the value, annually. This takes various costs into account, including the cost of money, space, handling, insurance, damage/degradation risk, and pilferage risk.

• What’s the organization’s required rate of return on an investment? Many have a “hurdle rate” of 10% to 15% for investments of capital, which is exactly what the purchase of a critical spare represents.

From there, we can calculate the internal rate of return (IRR) and net present value (NPV) for purchasing and managing a critical spare. In our example, assuming a 22% inventory-carrying cost and 10% required rate of return, a critical spare to protect a single system produces a -$76,000 NPV and a -15.3% IRR (a poor investment). If, however, that same spare can serve eight operating assets, the investment produces a $64,000 NPV and a 31.3% IRR (a wise investment, with the proviso you’ve factored in the conditional probability that more than one asset will demand the same spare at the same time). I’ll provide some worked examples of the process in an upcoming e-newsletter article.

In the end, decisions to store critical spares always involve some guesswork. With a little economic analysis, we can improve them. EP

Based in Tulsa, OK, industry veteran Drew Troyer is principal with Sigma Reliability Solutions. Email Drew.Troyer@sigma-reliability.com.