BUILDING MANAGEMENT SYSTEMS


Machine learning leads to enhanced BMS performance in education


George Catto, client services director for AMR DNA, a machine learning energy analytics service from Energy Assets, discusses how machine learning can enhance BMS performance in education


A


startling statistic in the government policy paper ‘Sustainability and climate change: a strategy for the education and children’s services systems’ - estimates


that schools and universities are responsible for 36% of total UK public sector building carbon emissions. So, no wonder the Department for Education’s


recently updated sustainability and climate change strategy for England highlights the contribution that building adaptation and decarbonisation across this estate could make to Net Zero.


A key aim of the strategy is to reduce


direct and indirect emissions from education premises, but with many establishments already under financial and resource pressure, how can this be achieved? With the right investment, this could in part be delivered by the installation of renewable power generation and storage. Indeed, it is the ambition of the UK government’s Solar Taskforce to include education within plans to increase solar capacity to 70GW by 2035 to power up Britain using cleaner, cheaper and more secure energy sources. While this relies heavily on the installation of solar PV roof and ground arrays on private premises such as warehouses and supermarkets, it also sees a significant role for publicly owned buildings, including schools and colleges. Even then, though, any such investment needs to be accompanied by processes integrated within building management systems (BMS) that optimise the value of clean energy by eradicating energy waste…which experience


shows can often be hiding in plain sight. Monitoring consumption and learning the lessons of waste


The good news is that just such an energy waste eradication programme is already underway in the higher (HE) and further (FE) education sectors, thanks to the foresight of The Energy Consortium (TEC).


TEC is a Contracting Authority owned by its members which delivers a wide range of services in energy procurement, data reporting, risk management and cost reduction on a not- for-profit basis.


TEC is working with AMR DNA, an Energy Assets machine learning and artificial intelligence data analytics service, powered by kWIQly, to improve energy efficiency across a number of university campuses.


In 2023 alone, this partnership identified and stopped 101 significant energy waste events with a notional value of £345,000. In addition, 14 new non-waste KPIs, tracking measures such as high summer base loads, poor timeclock control and overcompensation for weather variation, were incorporated by AMR DNA into BMS strategies, resulting in a further £325,000 of waste addressed. As a result, campuses adopting this AI-


informed machine learning approach have seen a 30% decline in the average duration of major energy waste incidents and a 60% reduction in total energy waste per month over the last two years. Of course, energy monitoring and reporting platforms that assimilate half hourly data from automated meter reading systems are not new. Systems such as WebAnalyser from Energy Assets provide a convenient way of comparing actual consumption versus benchmark parameters, enabling managers to measure the impact of efficiency programmes. However, extracting maximum value pre- supposes that the underpinning energy performance profiles for each building possess the necessary accuracy to act as benchmarks. It cannot be assumed that this is the case for every educational establishment. This is why machine learning and AI are fast emerging as a favoured route to energy efficiency and energy waste eradication across the HE and FE sectors. AMR DNA can crunch years’ worth of metered energy data in short order and progressively ‘learn’ what optimal performance looks like for each building. The system then applies pattern recognition to consumption models to spot tell-tale signs of energy waste unique to each building. These events can result from something as simple as equipment running needlessly or heating controls being incorrectly set. As such, machine learning can be a complementary tool to traditional monitoring and reporting platforms.


Illustrations:


These charts show anonymised data for a university site, illustrating the power of machine learning to identify and help eradicate energy waste and influence consumption behaviour. The graph above shows the impact of machine learning when it was applied to a university building heating system in May/June last year. Timeclocks and set points were altered and from that point on the consumption on Saturdays and Sundays, shown on the y-axis in kW, is significantly lower. The lower section shows the accumulative deviation from the model. on the left, the chart shows the two


different modelled consumptions patterns against temperature before and after the changes made on site. The “after” shows the model’s new expected consumption against temperature, with the site now using 100kW less at some temperatures. Note also before and after Saturday and Sunday lines (brown and gold), which show significantly lower consumption than before, due to shorter heating periods on a Saturday and turns off on a Sunday.


Developing an energy consumption hierarchy


So, you’ve got the data, where next? Machine learning tools can not only identify energy waste, they can also help create a consumption hierarchy to prioritise the actions that will deliver the fastest payback on cost or carbon reduction.


It’s worth the effort, because government


research suggests that a reduction of just 1% in average heating temperatures can lead to an 8% cost saving.


Any successful plan will likely include a mix of practical steps, such as adjusting and tailoring heating to match occupancy levels, installing low-energy lighting etc, along with behavioural change. One Energy Assets customer uses operational data to drive its ESOS actions and measure the contribution to efficiency of its investment in renewable energy. At the same time, staff engagement around energy efficiency is encouraged through a culture of shared ownership and individual responsibility. This includes nominating an Energy Champion to undertake a daily energy walk to help to eradicate waste, identify inefficient equipment usage and flag poor energy habits. And, of course, in the case of education, the bonus could be a transformational learning opportunity for young people witnessing at first hand energy efficiency and sustainability gains brought to life through advanced technology.


12 BUILDING SERVICES & ENVIRONMENTAL ENGINEER JULY 2024


Read the latest at: www.bsee.co.uk


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42