Summary: where to start cutting sag downtime
To cut downtime from voltage dips, facility teams should do six things in order: 1) compile past trips into a sag incident list; 2) calculate the actual loss from each shutdown; 3) set priorities by "number of trips × loss per trip"; 4) start with the lowest-cost improvements; 5) run a POC on the tool that trips most, keeping a control group; 6) once you have data, scale up and write sag ride-through into purchasing specs for new tools.
Step 1: Compile past trips into a "sag incident list"
You can start without power quality measurements. Sags leave traces in many places; line up the timestamps in these records and you have a credible incident list:
| Source | What you'll find |
|---|---|
| Alarm logs | Exact trip time and alarm type (power fault, servo alarm, EMO) |
| Downtime reports | How long it was down, output affected, any scrap |
| Work orders, shift logs | Restart time, who, what |
| Power monitoring | May have recorded the sag |
| Existing UPS event logs | Input anomalies confirm a sag |
| Utility and weather data | Ask the utility; check lightning records |
Add interviews with on-site staff: what they saw, which tools tripped, how long until production resumed.
Step 2: Cost each shutdown
Management wants numbers. The loss from one sag-related shutdown breaks down into three parts:
- Downtime=tools × recovery hours × hourly output
- WIP=reworked or scrapped WIP
- Recovery=requalification and labor
The three together give the "loss per incident"; multiply by the number of incidents per year for your annual sag risk. You can plug your numbers straight into the downtime loss calculator on the product page. Label estimates as estimates; don't present them as measurements.
Step 3: Set priorities by "trips × loss per trip"
You don't need to protect the whole plant at once. Multiply each model's annual trips by its loss per trip, rank the top few, and start with the biggest losses. You'll usually find the losses concentrated in a few models, often the same model tripping together.
Step 4: Cheapest fixes first
From lowest to highest cost: first check with the equipment maker whether undervoltage relay and VFD protection settings can be adjusted; then harden components such as contactors and control power supplies; next add voltage sag compensators to the control circuits that reset; use an online UPS only for loads that can't go down even in an outage. For when each method fits and its limits, see "How to prevent voltage sag trips in a factory"; for upgrading components on older tools, see "Older equipment fails SEMI F47. How do you improve it?"
Step 5: POC with a control
Improve the tool that trips most first, and use an unimproved tool of the same model as the control. The next time a sag hits, the results on both sides are the most convincing evidence.
Note: "no trips after the improvement" only means something if a sag actually occurred during that period. So monitoring is essential: a power quality analyzer or the sag compensator's own event log should record the time, depth, and waveform of every sag.
Step 6: Scale up and specify it
Once the POC proves effective, replicate it on the same model. More importantly, go one step further: write sag ride-through into purchasing specs for new tools, for example by requiring SEMI F47 compliance and reserving compensator connection points on control circuits. That way the problem doesn't start over every time a new tool arrives. For F47 requirements and how to read test reports, see "What is SEMI F47."
The one-page proposal
Sags last year, tools hit
From the incident list in Step 1, with data sources.
Loss per event, yearly risk
From the calculation in Step 2, with estimates labeled as estimates.
Fix and cost
Which tools first, which methods, and how much.
Payback condition
If the cost is less than one incident's loss, it pays for itself at the first sag.
Verification
POC tool, control group, and monitoring method, explaining how effectiveness will be proven.
FAQ
I'm on the facilities team and want to cut downtime from voltage dips. What's the first step?
The first step is to compile past trips into a sag incident list: cross-check timestamps from equipment alarms, downtime reports, work orders, UPS event logs, and facility power monitoring to find which trips were caused by voltage sags and which tools were affected. With that list, you can calculate losses and set priorities.
Can we start improving without power quality measurement data?
Yes. Equipment alarm logs, downtime reports, and UPS input event logs all leave traces of sags; line up the times to confirm them. Then add monitoring to capture data going forward.
How do I convince management to budget for sag improvements?
Let the numbers speak: list the past year's sag-related trips, the loss per incident, and the annual risk, then compare them with the cost of the fix. If the cost is less than one incident's loss, it pays for itself at the first sag. Prove the results with a one-tool POC first, then request a larger budget; that has the highest success rate.
How do you prove the improvement actually works?
Keep an unimproved tool of the same model as a control group, and record every sag with a power quality analyzer or the compensator's event log. If a sag actually occurred during the period, and the improved tool didn't trip while the control did, that's the most direct evidence.