- Advanced
- About 17 min
- 5 practice questions
Quantifying risk: expected loss, design life and cost-benefit
Calculate expected annual loss, judge mitigation by benefit-cost ratio, and find the chance of exceeding a design event during a structure's life.
If something looks wrong, please tell us.
At the advanced level, risk stops being a word and becomes a number that can guide spending. Planners ask how much a hazard costs on average each year, how likely a design flood is to be exceeded during a structure's life, and whether a mitigation project returns more than it costs. This lesson builds those three tools. All money figures are illustrative and kept undiscounted so the logic stays visible.
Expected annual loss
If a hazard event has an annual probability p and would cause a loss L, its contribution to the average yearly loss is p × L. Adding over all the events considered gives the expected annual loss (EAL): EAL = Σ (p × L). The EAL is an average over many years, not a forecast for next year; most years will see no loss and a few will see large ones. Rare, extreme events can dominate the EAL even though they almost never happen, because their losses are so large.
Worked example 1: EAL and a mitigation decision
A town faces two flood scenarios: a 10-year flood (p = 0.1) causing ₹20 lakh of damage, and a 50-year flood (p = 0.02) causing ₹100 lakh. Step 1: EAL = 0.1 × 20 + 0.02 × 100 = 2 + 2 = ₹4 lakh per year. Step 2: an embankment costing ₹15 lakh would halve the damage in both scenarios, so the new EAL is ₹2 lakh, a saving of ₹2 lakh per year. Step 3: over an assumed 10-year life without discounting, benefits = 2 × 10 = ₹20 lakh. Step 4: benefit-cost ratio = 20 ÷ 15 ≈ 1.33. A ratio above 1 means the measure returns more than it costs, so on these numbers it is worth building. A real appraisal would discount future benefits, which lowers the ratio, and would add benefits that are hard to price, such as lives saved.
Exceedance over a design life
Engineers design structures for an event of a chosen return period T. The chance that the design event is equalled or exceeded at least once in n years is 1 − (1 − 1/T)^n. For large T, a good shortcut is 1 − e^(−n/T). When n equals T, the chance is about 63 per cent, not 100 per cent and not 50 per cent. This is why critical structures such as dams and hospitals are designed for much rarer events than ordinary buildings.
Worked example 2: is the design safe enough?
A bridge is designed for a 50-year flood and expected to last 25 years. Step 1: n ÷ T = 25 ÷ 50 = 0.5. Step 2: shortcut chance = 1 − e^(−0.5) ≈ 1 − 0.607 = 0.39. Step 3: exact chance = 1 − 0.98^25 ≈ 1 − 0.603 = 0.40. So there is about a 40 per cent chance that the design flood is exceeded during the bridge's life. If the owner wants this chance below about 10 per cent, n ÷ T must be about 0.1, so T must be about 250 years.
| n ÷ T | Chance of at least one exceedance (about) |
|---|---|
| 0.1 | 10 per cent |
| 0.5 | 39 per cent |
| 1 | 63 per cent |
| 2 | 86 per cent |
Reading the factors together
These numbers sit on top of the conceptual relationship Risk = Hazard × Exposure × Vulnerability ÷ Capacity. The probability p describes the hazard; the loss L depends on exposure (what is in harm's way) and vulnerability (how much of it is damaged); capacity reduces L by enabling people to protect assets and recover. A mitigation measure can therefore work on any factor: zoning lowers exposure, retrofitting lowers vulnerability, and warning systems raise capacity.
EAL = Σ (probability × loss). Benefit-cost ratio = benefits ÷ costs, worthwhile above 1. Chance of exceeding a T-year event in n years = 1 − (1 − 1/T)^n ≈ 1 − e^(−n/T); about 63 per cent when n = T.
Key takeaways
Expected annual loss turns risk into an average yearly cost; benefit-cost ratios compare it with the cost of mitigation; and design-life exceedance shows why return periods must be much longer than the life of important structures. Always state assumptions such as discounting, and remember that unpriced benefits like lives saved strengthen the case for mitigation.
Check your understanding
-
1. A hazard has a 5 per cent annual chance of causing ₹40 lakh of loss and a 1 per cent annual chance of causing ₹300 lakh. What is the expected annual loss?
- ₹5 lakh
- ₹340 lakh
- ₹3.4 lakh
- ₹2 lakh
Show answer
Answer: A. EAL = 0.05 × 40 + 0.01 × 300 = 2 + 3 = ₹5 lakh per year.
-
2. A retrofit costing ₹30 lakh cuts expected annual loss by ₹4 lakh. Over a 15-year life, undiscounted, what is the benefit-cost ratio?
- 0.5
- 1.5
- 2.0
- 7.5
Show answer
Answer: C. Benefits = 4 × 15 = ₹60 lakh; BCR = 60 ÷ 30 = 2.0.
-
3. A structure is designed for a 100-year event and expected to last 100 years. About what is the chance the design event is exceeded at least once during its life?
- 100 per cent
- About 63 per cent
- 50 per cent
- 1 per cent
Show answer
Answer: B. With n = T, the chance is 1 − (0.99)^100 ≈ 1 − e^(−1) ≈ 0.63.
-
4. Banning new construction on a floodplain mainly reduces which factor of risk?
- Hazard frequency
- Magnitude of the flood
- Capacity
- Exposure
Show answer
Answer: D. Zoning keeps people and assets out of the flood's path, which reduces exposure; the flood itself is unchanged.
-
5. Why can a very rare event dominate the expected annual loss?
- Rare events have high probabilities
- EAL counts only rare events
- EAL ignores frequent events
- Its probability is small but its loss is so large that p × L is still big
Show answer
Answer: D. EAL adds p × L for all events; a tiny p multiplied by an enormous L can exceed the contribution of frequent small events.
Sign in to keep track of the lessons you finish.