Why add a discount rate when simple payback is the industry norm?

Because they answer different questions. Simple payback asks how long the money is at risk, which is a legitimate screening question. NPV asks whether the project creates value at the cost of the capital it consumes, which is the question that gets it approved. On these defaults the two differ by nearly half the headline benefit, and a case presented on payback alone will be re-appraised on NPV by someone else anyway.

What discount rate should I use?

Whatever your organisation applies to capital projects, for most manufacturers a hurdle rate well above the cost of debt, commonly 8–15%. If you genuinely do not know, run the page at both ends of that range: if the project clears at 15% the rate is not the deciding factor, and if it fails at 8% no amount of argument about the rate will save it.

Where do I get the annual mercury savings figure?

Build it rather than estimate it. The mercury energy page annualises electricity; the lamp replacement page gives cost per operating hour, which annualises with your runtime; then add swap downtime, exhaust and make-up air, and the building cooling load the lamps create. Doing this usually moves the number substantially in one direction or the other, and it is the input every row on this page depends on.

Why does the NPV equal the net cash when I set the rate to zero?

Because that is what a zero discount rate means: a dollar in year five counts as a dollar today, so the present value of five equal payments is simply five times one of them. It is worth doing once. It makes visible the assumption that undiscounted payback carries silently, and it confirms the two rows are computing the same cash flows.

What does this page leave out that could change the answer?

Three things, any of which can dominate. Tax, depreciation and utility rebates, which are jurisdiction-specific and can be large. Throughput: if the retrofit lets the line run faster or with fewer stops, the contribution from extra output usually exceeds every energy saving on the page. And reformulation, if the existing chemistry will not cure under the proposed array, which belongs in installed cost and is frequently the biggest single line in it.

Do UV LED arrays need replacing?

Eventually, yes. They degrade gradually rather than failing, and their lives are measured in tens of thousands of hours rather than thousands, but the end of that life is a capital-scale purchase, not a consumable. A maintenance figure that leaves out a provision for it makes the LED side look permanently cheap and quietly biases every row on this page.

UV Curing calculator

UV LED Payback Calculator: Simple Payback and Discounted NPV

Appraise a UV LED retrofit on both the basis everyone quotes and the basis a finance function will ask for. Enter the installed cost, the annual mercury cost the retrofit removes, the ongoing LED cost, and your discount rate. The calculator returns simple payback, net annual savings, five-year net cash and ROI, all as the earlier version computed them. Plus the five-year NPV of the same cash flows once the cost of capital is applied. Set the rate to zero and the NPV row collapses onto the net cash figure, which shows exactly what an undiscounted payback was assuming.

What this calculator does

Appraise a UV LED retrofit on both bases: the simple payback everyone quotes, and the discounted value of the same cash flows once a cost of capital is applied.
Use it for screening a UV LED retrofit before committing engineering time, presenting a capital case that will face a hurdle-rate test, showing how much of a payback claim is undiscounted arithmetic, testing sensitivity to the savings estimate, comparing retrofit options with different installed costs and running costs.
Appraise a UV LED retrofit on both bases: the simple payback everyone quotes, and the discounted value of the same cash flows once a cost of capital is applied.

Formula used

Net annual savings = annual mercury savings − annual LED maintenance
Years to payback = retrofit installed cost ÷ net annual savings
Five-year net cash = net annual savings × 5 − installed cost
Five-year ROI (%) = five-year net cash ÷ installed cost × 100
Five-year NPV = net annual savings × (1 − (1 + r)^−5) ÷ r − installed cost

Inputs explained

Retrofit installed cost: Arrays, drivers, cooling, controls, mechanical integration, electrical work and commissioning: everything spent before the first good part.
Annual mercury savings: Annual cost the retrofit removes: energy, lamps, swap downtime, exhaust and cooling. Build it from the energy and lamp-replacement pages rather than estimating it. This input decides the whole answer.
Annual LED maintenance: Ongoing LED cost: array electricity, cooling maintenance, and a provision for eventual array replacement, which is a capital-scale event rather than a consumable.
Discount rate: Your cost of capital or hurdle rate. Set it to zero to reproduce the earlier version's undiscounted figures exactly, which is what the old page was assuming without saying so.

How to use the result

Best suited to screening a UV LED retrofit before committing engineering time, presenting a capital case that will face a hurdle-rate test, showing how much of a payback claim is undiscounted arithmetic, testing sensitivity to the savings estimate, comparing retrofit options with different installed costs and running costs.
Assumes level savings. Real ones drift with electricity prices, production volume and lamp prices. Ignores tax, depreciation, grants and utility rebates, all of which can move a UV LED case materially. Prices no throughput change. If the retrofit lets the line run faster, that benefit dwarfs the energy line and is not on this page. Assumes the existing chemistry cures under the proposed array. Where it does not, reformulation cost belongs in the installed cost and is often the largest single item. No residual value on either side, and no allowance for array degradation raising LED energy over the horizon.

Current U.S. benchmarks

As of 2026-08-20, the U.S. prime lending rate is 6.75% (Federal Reserve via FRED). Equipment loans and lines of credit typically price at prime plus a spread, so use your actual borrowing rate when you have it.

Common questions

Why add a discount rate when simple payback is the industry norm? Because they answer different questions. Simple payback asks how long the money is at risk, which is a legitimate screening question. NPV asks whether the project creates value at the cost of the capital it consumes, which is the question that gets it approved. On these defaults the two differ by nearly half the headline benefit, and a case presented on payback alone will be re-appraised on NPV by someone else anyway.
What discount rate should I use? Whatever your organisation applies to capital projects, for most manufacturers a hurdle rate well above the cost of debt, commonly 8–15%. If you genuinely do not know, run the page at both ends of that range: if the project clears at 15% the rate is not the deciding factor, and if it fails at 8% no amount of argument about the rate will save it.
Where do I get the annual mercury savings figure? Build it rather than estimate it. The mercury energy page annualises electricity; the lamp replacement page gives cost per operating hour, which annualises with your runtime; then add swap downtime, exhaust and make-up air, and the building cooling load the lamps create. Doing this usually moves the number substantially in one direction or the other, and it is the input every row on this page depends on.
Why does the NPV equal the net cash when I set the rate to zero? Because that is what a zero discount rate means: a dollar in year five counts as a dollar today, so the present value of five equal payments is simply five times one of them. It is worth doing once. It makes visible the assumption that undiscounted payback carries silently, and it confirms the two rows are computing the same cash flows.
What does this page leave out that could change the answer? Three things, any of which can dominate. Tax, depreciation and utility rebates, which are jurisdiction-specific and can be large. Throughput: if the retrofit lets the line run faster or with fewer stops, the contribution from extra output usually exceeds every energy saving on the page. And reformulation, if the existing chemistry will not cure under the proposed array, which belongs in installed cost and is frequently the biggest single line in it.

Last reviewed 2026-08-25.