What published studies say about commissioning existing buildings
The North American studies on recommissioning, monitoring-based commissioning and fault detection, a UK case study, and how to use the figures honestly.
Anyone selling BMS optimisation is asked the same question sooner or later: what will it save? The honest answer is that it depends on the building. But there is published evidence on what commissioning existing buildings has achieved elsewhere, and it is worth knowing - and quoting carefully.
The main studies
- Crowe, Mills and others, Lawrence Berkeley National Laboratory (LBNL), 2020 - the largest dataset, nearly 1,500 buildings. Existing-building commissioning gave a median whole-building saving of about 6% (the middle half of projects saved between about 3% and 12%), with a median payback of 1.7 years. By route: 5% in utility programmes, 9% in monitoring-based programmes and 14% outside programme
- What got fixed - in the same study, 95% of the measures were schedules, setpoints, resets, sequences and general control fixes, and 87% were HVAC. That is the everyday work of a controls integrator.
- US Department of Energy Smart Energy Analytics Campaign, 2019-2020 - organisations using fault detection and diagnostics software saw median savings of about 9%, against about 3-4% for meter-level energy information systems, with paybacks of one to two years.
- Pacific Northwest National Laboratory (PNNL), 2021 - BMS trend data from 151 buildings showed a median savings potential of 12%, ranging from zero to 50%. PNNL's published re-tuning work reports savings of 5-20%.
Read them carefully
All of these are North American. The samples are often public-sector or university buildings, and the figures differ between studies because the samples, programmes and methods differ. An older and often-quoted LBNL figure of 16% has been superseded by the larger 2020 dataset. None of these studies predicts what any one building will save.
UK evidence is thinner but points the same way. A case study in CIBSE Journal (November 2023) described a 1960s multi-let office in Croydon where controls changes alone were reported to save 28% of its energy in nine months, verified under IPMVP - one building, not a benchmark.
The consistent message is more useful than any single number: most existing buildings have recoverable waste, finding it is skilled work, and the savings come from acting on what is found - and keeping it found.
How to use them
Say "published studies have found", name the source, and pair it with a plan to measure the result on the client's own building - a baseline, degree-day normalisation and a CUSUM chart. Our article Proving savings explains how.