The cost of ownership (CoO) of an ESC is the true per-wafer cost of holding a wafer. It is calculated by spreading the ESC’s purchase price across every wafer it processes over its full service life. This number decides whether a worn ESC should be repaired, replaced with a new OEM part or second-sourced. It almost always gives a different answer from the purchase price alone. This article sets out the CoO formula and explains every input an equipment engineer needs to gather. It then works through an illustrative example. The example compares a new-only ESC against one that is repaired across multiple lifetimes.
It supports the cluster pillar on ESC repair and refurbishment. The pillar sets out the wider case for lowering cost of ownership on legacy 150 mm and 200 mm tools.
The cost of ownership formula
Cost of ownership for an ESC is the purchase price divided by the total wafers the ESC will process across its service life. It is expressed per wafer:
CoO = purchase price ÷ (service lifetime × repair cycles)
Service lifetime is the number of wafers an ESC processes between one recondition and the next. Repair cycles is the number of usable lifetimes you get from a single ESC body. That means one lifetime as new, plus each lifetime recovered by refurbishment. The principle is simple. The ESC body is the expensive part. Each repair recovers its clamping and particle performance without buying a new body. The one purchase is then spread across more wafers, and the per-wafer cost falls. An ESC that is never repaired carries its entire purchase price against a single lifetime. An ESC refurbished three times carries the same body cost across four lifetimes, plus the smaller cost of the repairs.
To use the formula in practice you need four inputs. You need a price, either purchase or replacement. You need a service lifetime in wafers. You need the number of repair cycles the body will support. And you need the cost of each repair. The sections below explain each input and where the largest costs are.
Input 1: purchase or replacement price
The first input is what it costs to put a working ESC on the tool. For a new ESC from the original equipment maker, this is the list price. For legacy 150 mm and 200 mm tools, that price is typically high and the lead time is long. Sometimes the part is discontinued altogether, so no price exists at any lead time. A qualified OEM-equivalent replacement performs the same function and clamps to the same specification. It costs less and arrives sooner. This is why second-source supply matters most on the tools the OEM no longer prioritises. Enter the price of the route you would actually take into the CoO formula, not a theoretical figure. The whole calculation depends on it.
Input 2: lead time and downtime cost
The purchase price is rarely the largest cost. An ESC that takes months to arrive keeps a tool idle. Idle tool time on a production line costs far more than the ESC itself. Downtime cost is the lost production value while the tool waits for its part. It rises directly with lead time. So a long-lead new OEM ESC can carry a hidden cost of several times its list price. That cost appears when you count the wafers the tool did not run. This is not part of the core CoO formula. But it belongs in any fair repair-or-replace comparison, because it is usually the factor that decides the outcome. A refurbishment or an in-stock OEM-equivalent can return the tool to service in weeks rather than months. That alone can make it cheaper overall, even before the lower part price is counted.
When you build the comparison, attach an estimated downtime cost to each route’s lead time. Keep it visible next to the per-wafer CoO. An equipment engineer defends both numbers to a fab manager: the amortised part cost and the lead-time risk.
Input 3: service lifetime in wafers
Service lifetime is how many wafers an ESC processes before its performance drifts out of specification and it needs reconditioning. The lifetime ends when clamping force falls too far. It also ends when the dielectric surface erodes and particle counts rise. Worsening helium leak-back and degraded backside cooling end it too. So does residual clamping that makes dechuck unreliable. Each of these is a measurable symptom at the tool. Together they set the lifetime figure you put into the formula.
Lifetime is process-specific. An aggressive plasma etch erodes the clamping surface faster than a gentler deposition step. So the same ESC design will show a different wafer count between recommended recondition points on different tools. The best source for this number is your own maintenance history, not a generic figure. Use the wafer count between past ESC swaps on that tool. Tracking clamping force and particle data over time shows where the real end-of-life point is. It also shows how much lifetime a recondition recovers.
Input 4: number of repair cycles
Repair cycles is the multiplier that lowers CoO. An ESC can typically be refurbished more than three times before the body reaches genuine end of life. Each cycle restores as-new clamping and particle performance. This means a single ESC body can deliver four or more usable lifetimes: the original plus the refurbished ones. The condition is that the underlying ceramic and base assembly remain sound. The limit is reached when the dielectric can no longer be re-surfaced within flatness specification. It is also reached when electrode or feed-through degradation can no longer be reworked.
This is the input with the largest effect on cost of ownership. Doubling the repair cycles roughly halves the share of the body price carried by each lifetime. Whether a given ESC body will support its full count of cycles depends on its condition at inbound inspection. This is why the repair-or-replace decision always starts with the body itself, not the surface wear.
Input 5: repair cost at L1 and L2
Refurbishment comes at two levels. The cost of each is expressed as a percentage of the new list price. An L1 surface recondition cleans, re-laps and re-surfaces the clamping face, then re-tests it. It costs roughly 25% of new list price. It addresses surface wear, mesa erosion and the rising particle counts that come with them. An L2 full strip and rebuild reworks the dielectric and electrode and fully reassembles the ESC. It costs roughly 70% of new list price. It addresses deeper damage such as dielectric wear, electrical degradation or loss of flatness that an L1 cannot recover. Both restore the ESC to as-new condition, warranted, with a full functional test before return.
Which level a lifetime needs depends on the failure mode that ended it. Surface erosion and particle drift usually need L1. Loss of clamping force, helium leak-up or rising leakage current usually point to L2. Most ESCs have several L1 cycles across their life, with an occasional L2 when deeper rework is needed. For the full scope of each level, see L1 vs L2 ESC repair. When you build the CoO sum, add the cost of each repair cycle to the original purchase price. The body cost plus the repairs is the total you are spreading across the wafers.
Worked example (illustrative numbers)
The figures below are illustrative examples chosen to show the method. They are not quoted prices. Replace them with your own price, lifetime and process data. Assume a new ESC list price of £20,000 and a service lifetime of 50,000 wafers per cycle. The repair percentages (L1 ≈ 25%, L2 ≈ 70% of list) and the typical repair count (more than three cycles) are the C6 locked figures. The £20,000 and 50,000-wafer numbers are illustrative only.
New-only. The ESC is run for one lifetime and scrapped. Its entire £20,000 price is carried by 50,000 wafers:
CoO = £20,000 ÷ (50,000 × 1) = £0.40 per wafer
With repair. The same ESC body is refurbished three times. For example, three L1 reconditions at 25% of list cost £5,000 each. This gives four lifetimes from one body. The total is the original £20,000 plus three repairs at £5,000, which is £35,000. It is spread across 4 × 50,000 = 200,000 wafers:
CoO = £35,000 ÷ (50,000 × 4) = £0.175 per wafer
| Input (illustrative) | New-only | With repair (3× L1) |
|---|---|---|
| New list price | £20,000 | £20,000 |
| Repair cost added | £0 | 3 × £5,000 = £15,000 |
| Total amortised cost | £20,000 | £35,000 |
| Usable lifetimes | 1 | 4 |
| Wafers per lifetime | 50,000 | 50,000 |
| Total wafers processed | 50,000 | 200,000 |
| Cost of ownership | £0.40 / wafer | £0.175 / wafer |
In this illustrative case, repairing the ESC across three cycles more than halves the per-wafer cost of ownership, even after paying for the repairs. The reason is that the expensive ESC body is spread over four lifetimes instead of one. Replace one of the cycles with an L2 rebuild and the total amortised cost rises. The per-wafer figure still stays well below the new-only number. Four lifetimes from one body is the dominant effect.
How repair and OEM-equivalent supply lower CoO
Two factors lower cost of ownership, and they combine. The first is repair. Each refurbishment cycle recovers a full lifetime for a small share of the new price. So the denominator in the formula grows much faster than the numerator. The second is the starting price you pay for the body. A new OEM ESC has the highest purchase price and, on legacy tools, the longest lead time. An OEM-equivalent replacement lowers that starting price and shortens the lead time. This reduces the per-wafer figure and also reduces the downtime cost.
Together, the two factors change the economics of running a legacy fab. Consider an ESC second-sourced at a lower price and then refurbished across several lifetimes. Its per-wafer cost is far below that of a new OEM ESC run once and scrapped. One condition makes this work: the refurbishment must genuinely restore as-new clamping and particle performance, proven by a full functional test. Otherwise the recovered lifetimes are shorter and the saving shrinks. To see where the repair-or-replace boundary falls in practice, see our repair or replace decision framework.
Frequently asked questions
How do I calculate ESC cost of ownership?
Divide the purchase price by the total wafers the ESC will process. That total is its service lifetime in wafers multiplied by the number of usable lifetimes. Each refurbishment recovers one lifetime. The result is a per-wafer cost. To compare repair against new, add the cost of each repair cycle to the original purchase price before dividing.
How many times can an ESC be repaired?
An ESC can typically be refurbished more than three times before the body reaches end of life. Each cycle restores as-new clamping and particle performance. The limit is set by the condition of the ceramic dielectric and base assembly. This condition is assessed at inbound inspection.
How much does it cost to repair an ESC?
An L1 surface recondition costs roughly 25% of the new list price. An L2 full strip and rebuild costs roughly 70%. Both return the ESC to as-new condition with a full functional test. Even an L2 costs much less than a new OEM ESC and recovers a full lifetime.