Is my site within normal service conditions?

1. Choose the site. Search for a place in the list included in the page, or paste a map link or coordinates. As soon as you choose the site, the page sends its position to NASA POWER to get the temperatures: only the position, rounded to 0.1° (about 10 km), without credentials, cookies or referrer. No transformer data is sent. The climate always comes from NASA POWER, daily maximum and minimum temperatures for the last 20 complete years, today 2006-2025.
2. Add the transformer. Choose the paper (kraft or thermally upgraded), the cooling mode and, if you have it, the usual daily peak load. For capacity and the hour-by-hour results, enter the top-oil rise, the hot-spot gradient and the loss ratio from the heat-run test report (IEC 60076-2:2011). The tool does not assume them: without them it shows only the rated-load reference.
3. Read the dossier. Six parts: the IEC design basis today and in 2050, with a monthly chart; the weighted ambient temperature θE and the relative ageing rate of the paper; the continuous load in the hottest month for hot-spot limits of 98, 110 and 120 °C; a month-by-hour map of the hot spot; the hottest three-day spell in the NASA record; and the notes, defaults and sources.

The 2050 scenario defaults to SSP2-4.5, and you can choose SSP1-2.6 or SSP5-8.5. You can save the inputs to a file and open them later; Seetalabs® keeps nothing.

Design basis. IEC 60076-1:2011 clause 4.2 defines the normal service conditions: an ambient temperature not above 40 °C, a monthly mean of the hottest month not above 30 °C, a yearly mean not above 20 °C, and an altitude up to 1 000 m. The monthly means follow IEC 60076-1:2011 clause 3.12, the average of the daily maximum and minimum. Above these values, IEC 60076-2:2011 clause 6.3.2 reduces the temperature rise limits by the excess, rounded to the nearest kelvin, and by 1 K for every 400 m (natural air) or 250 m (forced air) above 1 000 m.

Ageing. The weighted ambient temperature θE follows IEC 60076-7:2018 clause 8.3.1, equation 27, and the relative ageing rate uses equation 2 for kraft paper and equation 3 for thermally upgraded paper. With a load, twelve typical days are simulated with the difference equations of IEC 60076-7:2018 clause 8.2.3.

Summer capacity. The continuous load at which the hot spot reaches the limit, with the mean of the hottest month as ambient (IEC 60076-7:2018 clause 8.3.1 b).

2050. A Seetalabs® calculation with the delta method: the monthly change in mean temperature of the CMIP6 models for the site's IPCC AR6 land region, 2041-2060 against 1995-2014, is added to the site's monthly means, with the median as the central case and the 10th to 90th percentile as the range. The regional data come from the IPCC WGI AR6 Atlas repository. The resulting 2050 values are our computation, not values published by the IPCC. Because the NASA record already contains part of the warming since 1995-2014, that part is removed by linear interpolation between the centres of the periods: for 2006-2025 about a quarter of the change.

The tool's reference test case is Rome, with the monthly means of the NASA POWER daily record 2001-2020 (hottest month 26.47 °C in August, highest temperature 40.92 °C, altitude 198 m), an ONAN transformer with kraft paper, a top-oil rise of 52 K, a hot-spot gradient of 26 K and a loss ratio of 6.

The site is outside the normal conditions on one criterion only: the highest recorded temperature is 0.92 K above 40 °C. Rounded to the kelvin, the correction is 1 K, so a new unit for this site would be specified with rise limits of 59, 64 and 77 K instead of 60, 65 and 78 K. A standard unit already in service is not wrong, but it ages faster and carries less in summer: 0.937 p.u. continuous at a 98 °C hot spot today, 0.917 p.u. with the 2050 climate, about 2 % less.

This test case applies the full regional change, because its record is 2001-2020. When you choose Rome in the tool, it downloads the latest 20 complete years and removes the part of the change already in them, so the numbers you see will differ slightly.