Technical guideFree to read2014

Demystifying Generator Set Ratings

Caterpillar · Ray Marfell

Caterpillar LEXE0549-01 (Ray Marfell, 2014) explains ISO 8528-1 rating classes and how Cat's own ratings map onto them. It argues that data centres are often over-specified by designing for constant full load at worst-case ambient.

On this page

Overview

The paper treats a rating as an application permission for the same engine iron. Cat's ratings (ESP, Standby, Mission Critical Standby, Prime, Continuous, Load Management) map to ISO classes. ISO Prime allows unlimited hours at an average load factor of up to 70% of the Prime rating, with a 10% overload for at most 1 h in 12 and at most 25 h per year. ISO LTP (Cat Load Management) uses the Prime nameplate, up to 100% average load factor, at most 500 h per year and no overload.

ESP is limited to 200 h per year with a 24-hour average of no more than 70% of ESP. Cat Standby allows up to 500 h per year at no more than 70% average and is not for utility paralleling. Cat Mission Critical Standby, aimed at data centres, allows a higher average load factor of 85%, with time at nameplate limited to 5%.

On real load profiles, the author shows field data from 2.5 MW-rated sets at data-centre sites, including commissioning tests, that fits within the Standby rating. UPS battery recharge after an outage typically adds 5-10% load for a few minutes, while an under-loaded UPS raises harmonics and calls for a larger generator end. The paper closes by warning against specifications that quote several ratings for one machine.

Key points

  1. ISO Prime: unlimited hours, average load factor at most 70%, 10% overload for 1 h in 12 and at most 25 h/year.
  2. ISO LTP (Cat Load Management): Prime nameplate, up to 100% average load factor, 500 h/year maximum, no overload.
  3. ESP: 200 h/year maximum, 24-hour average at most 70% of ESP.
  4. Cat Standby: 70% average, up to 500 h/year, not for utility paralleling.
  5. Mission Critical Standby: 85% average load factor, time at nameplate limited to 5%, aimed at data centres.
  6. UPS battery recharge adds 5-10% load for a few minutes after an outage.
  7. Higher-hour applications typically follow a 500-hour or annual oil-change interval, so 'unlimited' running still needs maintenance planning.
  8. Field data from 2.5 MW sets at data-centre sites fits the Standby rating.

Using it in design

  • Store per rating the kW, average-load-factor limit, annual-hour limit, overload allowance and a utility-paralleling-permitted flag.
  • Check expected average load against the selected rating's limit: 70% (Standby/ESP/Prime), 85% (Mission Critical), 100% (COP/LTP).
  • Reject specifications that list multiple ratings for one set without naming the application.
  • Add a 5-10% UPS recharge transient to the load model on retransfer to generator.

Monthly newsletter · free

Keep your reading list current

Once a month: new library sources, changes to Turkish regulations and upcoming data centre events.

SoSo CAD

Review your power design in SoSo CAD

Inspect supply paths, equipment loads and UPS failure scenarios in one model.

Request a demo
Single-line diagram in the SoSo CAD power and capacity workspace

Who should read it

Data-centre electrical designers choosing generator ratings and writing specifications.

Related standards

  • ISO 8528-1
  • ISO 3046-1

Publication details

Authors
Ray Marfell
Publisher
Caterpillar
Year
2014
Document number
LEXE0549-01
Document language
English
Source type
Manufacturer
Access
Free to read
Licence
Copyright retained; citation only — the publisher’s terms apply to the original document

Citation

Ray Marfell (2014). Demystifying Generator Set Ratings. Caterpillar. LEXE0549-01. https://www.finning.com/content/dam/finning/en_gb/Documents/Industries/ElectricPower/cat-white-papers-march-2024/Demystifying%20Generator%20Set%20Ratings%20(LEXE0549-01).pdf

All content and rights belong to the respective authors and publishers. SoSo CAD provides only the citation, its own summary and a link to the source, and is not responsible for the sources' content.