It’s easy to read mean time to repair (MTTR) as the time spent repairing. In solar O&M, the field data tells a different story.
At a 46 MW plant in Northern Italy, researchers at Politecnico di Torino tracked three years of inverter maintenance records. Replacing a failed IGBT took about 95 hours on average while spare parts took up to 40 days to arrive. In 2017, with parts available, the same repair averaged about 16 hours (Spertino et al., 2021).
The work didn’t get faster. The waiting got shorter.
That’s the pattern to look for in your own MTTR: time spent waiting for someone to see the alarm, for the right part, or for a second trip because the first one was missing something. A fast response isn’t the same as a fast fix. What counts is fixing it on the first visit.
Why the waiting is getting harder to manage
Raptor Maps’ 2026 Global Solar Report, based on 373 GW of US utility-scale and C&I sites, puts average power loss at 5.08%. That’s worth up to $5,070 per MW a year in lost revenue (pv magazine, Feb 2026).
Two findings from the same report matter for repair times:
- Each technician now covers about 70% more MW than five years ago. Every unnecessary trip costs more of a stretched team.
- Losses are moving away from inverters. Strings (26.9%) and combiners (21.5%) now cause more of the loss than inverter faults. Those faults are smaller and spread across far more points on site.
The data covers US sites. The direction it points to is clear enough: more faults to find, in more places, with fewer people per MW.
Four ways to reduce MTTR in solar O&M
Each one removes a wait. Here’s how each works in ACTIS ERP.
1. Cut mean time to detect
Solar fault detection can’t be faster than the data behind it. ACTIS receives plant data from your monitoring platform every 15 minutes, so a fault shows up as an alert within about a quarter of an hour.
Speed is only half of it. Alerts can use weather-adjusted thresholds, and ACTIS deduplicates, suppresses and groups them. One grid event doesn’t fill the queue with duplicates, and the people on call see the faults that need action.
2. Remove the hand-offs between alert and dispatch
In ACTIS, an alert can open an incident automatically, and the incident can open a work order. Each client decides how much runs on its own. ACTIS suggests who to assign, with technician location and availability in view.
Every manual step you take out is one less place for a fault to sit unnoticed.
3. Fix it on the first visit
This is where the Torino hours went. When a technician adds parts to a work order, ACTIS shows stock levels and which warehouse or van holds them. Each part used is traced to its work order and incident.
On site, the ACTIS mobile app works offline. Photos, parts, time and task lists are logged even without signal, so nothing waits for the drive back. If a part has to be ordered, the purchase request and its approval can run through ACTIS’s procurement module, which can be connected to spare-parts stock.
4. Learn from every fault, and measure every step
Each incident records a root cause from a fixed list, plus free-text notes. Engineers can search past incidents by site, technician, fault or date range, so a repeat fault starts from what fixed it last time.
Every step is timestamped: alert, acknowledgement, assignment, arrival, resolution, closure. ACTIS’s standard Power BI reports split MTTR into those intervals. Lost energy and cost per incident are calculated automatically and appear in the monthly report. You see which wait is costing you most, instead of one average.
You run a fleet, not a call centre. Measuring the response time alone tells you how fast you answered, not how fast you fixed it.
See the details on our Service Management page, or how ACTIS works for O&M service providers.
Built by people who run plants
ACTIS was built by Alectris, a solar O&M and EPC company, and Alectris has run its own operations on it for more than ten years. Its field process runs on ACTIS today.
In August 2026 Alectris published what 16 months of its own categorised work orders showed. Across the O&M operations it manages, the median intervention time between January 2025 and April 2026 was 137 minutes (Alectris, Aug 2026). That figure is an O&M process result, measured in ACTIS, not a software benchmark.
Five questions to ask about your own MTTR
- How long does it take for a fault at the plant to become an alert that someone acts on?
- How many manual steps sit between that alert and a technician’s work order?
- When a technician leaves for site, do they know the part they need is in stock, and where?
- How often does a fault need a second visit, and do you record why?
- Can you split your MTTR into detection, response and repair time, per site?
If your current tools can’t answer two or more of these, the waits are hidden inside your average.
FAQ
What’s the difference between mean time to respond and mean time to resolve?
Mean time to respond measures how long it takes to get someone to the fault. SolarPower Europe’s O&M Best Practice Guidelines define response time as acknowledgement plus intervention, up to the technician reaching the plant. Mean time to resolve, or repair, continues until the fault is fixed. A contract can meet its response time while the plant stays down.
How do you calculate MTTR for a solar plant?
MTTR is the total downtime of all corrective events in a period, divided by the number of events:
MTTR = total corrective downtime ÷ number of corrective events
Ten faults with 60 hours of total downtime give an MTTR of 60 ÷ 10 = 6 hours. Fix the start point (alarm or acknowledgement) and the end point (fault resolved) once, and use them for every site, or the numbers won’t compare.
What is a good MTTR in solar O&M?
There’s no single benchmark. It depends on the component, site access and the response times in your contracts. In the Torino study, inverter repairs averaged about 9 hours overall, and one component went from about 95 hours to about 16 once parts were available. The more useful target is your own trend, split into detection, response and repair.
See how it works live
In a demo we’ll follow one fault through ACTIS, from the first alert to the closed work order and the report.
