LOW CARBON BUILDINGS
Can machine learning & AI transform energy efficiency?
A
There’s probably never been a more important time for industrial and commercial (I&C) organisations to get ‘energy fit’ as part of their wider sustainability goals, says David Sing, managing director, metering and data, Energy Assets
lthough the cost of energy has reduced somewhat since prices peaked, I&C organisations
are still facing historically high bills. As a result, businesses are more aware today than ever before of the need to bear down on consumption while reducing their carbon footprint. Consequently, there’s growing
interest in the role technology can play in improving building energy efficiency. And when we say technology, we mean machine learning and artificial intelligence (AI) in particular.
AI is constantly in the headlines, but the question we’re being asked at Energy Assets is whether AI and machine learning can really make a difference to energy consumption. Many are surprised to learn that it’s already here…and it’s helping organisations drive out energy waste. Of course, AI-informed machine
learning can only be as good as the data that feeds it. So before companies look to AI to improve their energy profile, there are some basic steps to be taken to
lay the foundations for machine learning. This includes making sure that consumption data is being captured in granular detail via automated meter reading systems. Many businesses will already employ advanced AM&T portals, such as our own WebAnalyser, to monitor and report on consumption, run bespoke reports, and set automated alerts for unusual spikes in consumption. These may provide sufficient analysis for many users, but the advance of AI and machine learning certainly provides an opportunity to go further.
Heavy data lifting
At Energy Assets, we recognised the potential impact of AI and machine learning a few years ago and launched AMR DNA, powered by kWIQly, on the back of the availability of hour-hourly metered data. What’s different about machine
learning is its ability to learn what optimal performance looks like in a single building or across an entire portfolio. It can progressively assimilate meter data and quickly produce a checklist of actions to
The Royal College of Music adopts machine learning to target Net Zero
One of the world’s foremost institutes for musical excellence is adopting machine learning to help orchestrate sustainable building performance. The Royal College of Music (RCM), ranked the No. 1 global institution for performing arts, is harnessing the power of artificial intelligence (AI) to bring energy efficiency into harmony with plans to achieve Net Zero carbon by 2035.
The RCM estate includes a patchwork of buildings ranging from ultra- modern performance and teaching halls to Grade II listed premises dating from the 19th Century. Its campus has been transformed in recent years, with a near doubling of the estate’s footprint, thanks to the addition of a new performance hall, a new performance studio, a large café and courtyard area, and a new interactive museum. All contribute to an inspirational environment in which students from around the world create, research and perform.
Fine-tuning energy performance for Net Zero
To support its Net Zero carbon journey, the institution is now employing AMR DNA machine learning from Energy Assets across its Prince Consort Road site, which comprises six buildings of varying configuration and age.
The tool will be analysing half hourly meter data from a new network of electricity sub-meters to create consumption patterns that can spot tell-tale signs of energy waste, such as equipment running needlessly, or heating controls incorrectly set.
help energy managers implement efficiency strategies. And unlike us mere mortals, it never sleeps. Day in, day out, this tool is automatically gathering meter data and interpreting it in the context of building function and external factors (weather, occupancy levels). This produces ‘fingerprints’ that identify patterns of waste while ignoring outcomes that are irrelevant, mistaken or due to bad data.
Crunching data on this scale manually, particularly across multi- building portfolios, would require an army of analysts – but with machine learning, this can be achieved in a fraction of the time. It’s often a question of spotting
improvement opportunities hiding in plain sight, but sometimes these can be the hardest to identify because they are ingrained in a building’s legacy performance. AMR DNA is the perfect tool to spot these patterns of waste. Moreover, the value of machine
learning will only grow with the Market-wide Half-Hourly settlement (MWHH) reform due by 2025. This will not only underpin machine learning tools through the application of half-hourly data, it
will also enhance opportunities for I&C users to negotiate demand side response incentives and preferential time-of-use tariffs from their energy suppliers.
This AI-informed system will progressively learn what benchmark performance looks like, fine-tuning the effectiveness of the established building management system (BMS). AMR DNA already provides data analytics on gas consumption to the RCM via The Energy Consortium, a specialist procurer of low carbon energy services for higher education institutes. Now though, the RCM will be using the tool for its Prince Consort Road buildings to drill down deeper into
consumption routines and behaviours. This will deliver data in granular detail to support the institution’s broader Net Zero ambitions. “We have set ourselves the challenge to become Net Zero carbon by 2035,” said Gethin Lewis, Estates Projects & Environmental Coordinator at RCM. “To date, we’ve made good progress and by 2020 we reduced our Scope 1 and 2 emissions (gas and electricity) by 60% over the Higher Education Funding Council for England 2004/5 baseline target of 42% by 2020, even while our estate grew.” This was achieved through actions such as upgrading lighting to LED
fixtures, optimising energy control via a BMS, improving building thermal insulation and using boilers more efficiently.
An ensemble for a sustainable future
“Now though, we need to go further,” said Gethin. “While much of the focus will be on reducing Scope 3 emissions generated largely by services procurement, travel and transport, there is also more to do to cut energy-based emissions. Here, using the 2004/05 baseline, we aim to reduce our gas and electricity carbon emissions by 73% by 2027 and by 100% by 2035.”
In the case of gas, this means moving to lower carbon heat sources including heat pumps and point-of-use water heaters. “For electricity, we’ll be looking at voltage optimisation, air conditioning system rationalisation, lighting replacement and IT infrastructure efficiencies. This includes adopting machine learning in Prince Consort Road to understand how AI can help us spot waste and inform actions leading to optimal energy performance.” The analytics engine inherent in AMR DNA from Energy Assets, powered by kWIQly, has several report modes to save energy managers’ time. These include establishing achievable targets by KPI and weather condition for each meter, forecasting future use and identifying where changes have taken place; and diagnosing priority and cause of savings opportunities.
8 BUILDING SERVICES & ENVIRONMENTAL ENGINEER SEPTEMBER 2023 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 |
Page 43 |
Page 44 |
Page 45 |
Page 46
