Monitoring Baseload Consumption Across Hundreds of Meters

A recent BBC news article  reported the findings of a study on the impact of renewables compared to energy efficiency on our electricity consumption since a 2005 peak. While renewables have reduced fossil fuel energy by 95 TWh, energy efficiency has reduced demand by 103 TWh.

Energy efficiency can often be improved through low-cost solutions, so this article’s focus is on the energy savings achievable through simple main meter monitoring using a single alert type to target energy waste. I ran an energy management cycle from energy alerts, to desktop and site investigations, business cases, managing resolutions, M&V of results, continual monitoring and re-benchmarking for continual improvement.

The project operation was based in Bristol with me providing every part of the alert cycle with a senior consulting engineer as an advisor and key account manager. The client, the National Estate and Facilities Manager, was introduced into current work streams, data management tools, analytics and energy project management processes quickly in order to provide the requirements for this project that included access to kWh per half hour (HH) data in the form of a day+1 file and site information such as addresses, floor area and site contact details. Locations with their details were created on an energy management platform with the HH data uploading via email each day where the configured alerts were generated.

The client in question has c.200 warehouse retail sites with the utilities being used for HVAC, lighting and small power only. Energy and facilities management across the whole portfolio is outsourced to third party providers with the FM provider’s work force c. 10 sites per a Facilities Manager. Operating hours vary, but the general principle is 6 hours on a Sunday with 12 hours every other day, winged by up to an hour of staff occupancy. Requirements for out-of-hours (OOH) consumption is minimal.

Project Aim

The goal of this project was to find and realise opportunities for energy savings through the kWh per half hour data from main utility meters.

Staged objectives for the project were:

  • Introduce historical and daily HH data into the energy management platform,
  • benchmark the current HH profile for each site,
  • create and review alerts on the HH data,
  • investigate causes for high consumption leading to alerts,
  • manage the resolution of determined issues,
  • measure and verify the cost avoidance.

Due to the occupancy profile of the sites and the nature of the sites having minimal requirements for OOH consumption, the alert type discussed in this article was based on this period. High consumption during OOH may be as a result of a baseload increase, controls misuse such as overrides or operational faults by BMS or personnel. Other alerts for separate times of the day or for different measures were used in conjunction with this alert but this article will focus on the OOH alert.

From Benchmarking to Resolutions

Data received daily from the client’s data collector was sorted into the energy management platform where it was then available for detailed analysis and review. Benchmarks would be created from energy consumption patterns and knowledge of the typical asset specification and operation patterns. These benchmarks would be the foundation of our monitoring and targeting process.

The platform was set up to provide alerts based on these benchmarks, allowing for high or irregular energy consumption to be discovered through minimal data manipulation. In order to focus energy efficiency improvements, I concentrated investigations on recurring, significant issues i.e. 3 hours straight each night, every night for a week.

Once streamlined, investigations into the alerts would begin. Consisting of desktop investigations focussing around knowledge collection from onsite personnel such as site managers or facility managers around potential or known causes. In the instance of this project, I could also dial into all operational Building Management Systems (BMS) in order to review schedules, setpoints and control operations such as overrides and last man out controls.

Managing the resolutions, tracked through opportunity logs, would vary substantially dependent on the opportunity. Actions to resolve an issue could range from monitoring an operational change, to organising site visits from specialist BMS control engineers alongside the client’s FM technicians in order to find the complete resolution between equipment and controls.

Measurement and verification of the resolution was important to ensure that the savings impact was realised and maintained. Over a designated time period of 3 months, the savings achieved were monitored to confirm the continuation of the result. If the consumption was to revert to previous patterns or change drastically, the process would return to the investigation stage.

The Importance of Accurate Data

The first consideration in order to increase the effectiveness of this project was to ensure the accuracy and completeness of the data. As I could have received and filtered hundreds of alerts, their initial origin was important and should not have been due to erroneous data. Vice versa, missing data may have led to missed issues. While data completeness was monitored separately through regular data checks, the alerts were filtered based on the number of consecutive alerts and those with a full week of alerts had the previous month’s data sense-checked on the energy management platform.

The accuracy of the benchmark was also an influencing factor on the effectiveness of this project. Too lenient and too few alerts may have been raised, too strict and too many alerts may also have been raised. Due to the targeting of this alert to non-trading consumption, the main concern was that the benchmark is too low or strict and there are too many alerts from sites with only the necessary overnight equipment running.

The main challenge when investigating these issues was that they occurred when the site was unoccupied. Some of the issues and remedies discovered were post-energy efficiency installation works, for example LED fitouts (including stand-alone lighting controls) and BMS installations. Complications arising from the completion of these projects were discovered through the energy management platform’s alert system. The BMS and energy profile were not able to directly indicate the cause of the higher consumption, therefore we were next relying on the site manager’s knowledge of their new energy systems and what was still on when they left the premises (despite following usual shut down procedures). To gain a more technical insight, we would question the site’s FM. However, their visibility of what is running overnight is limited due to their working hours and travel requirements.


Stakeholder engagement became important when investigating the issue through site personnel and changing their operational habits. An example is one site having sections of lighting overridden and staff were manually switching the override due to incorrect BMS settings, leading to these sections of lighting being left on overnight frequently. The staff were unaware that if the issue was raised, I would be able to alter the settings, eliminating the requirement for manual switching and reducing the risk of consumption OOH.

Stakeholder buy-in was also important when discovering an issue that requires a cost to discover and fix. At one site an emergency override had been used in the back of the warehouse. This subsequently overrides the BMS controls and requires resetting by a BMS engineer on site. Due to this learning, a 12-hour reset of the emergency override was implemented to avoid the situation again. If used, the BMS will be overridden for 12-hours before reverting without the need of the additional cost of a BMS engineer site visit.

The benchmark or site-specific historical data can be used to provide predicted savings, indicating the baseload that can be achieved and therefore the cost savings that can be achieved over time. Often the solution is very low cost, sometimes these examples will be used alongside the solutions that require a cost (rarely high) in order to demonstrate the positive impact the resolutions have across the portfolio.


After 16 weeks of the process, over £200k avoidance in annual electricity cost was achieved due to remedies discovered and realised. Few of these remedies required a site visit and most were resolvable via desktop actions. Including the cost of the analyst for the time worked and the cost of any site work required; the cost of running this project could have been paid for more than 20 times over in the first year.

What Next?

The future of this project is set to continue, alongside other alerts and other measures being used to manage the portfolio’s energy and increase efficiency. From analysis into the portfolio in comparison to the current benchmark, a further £105k+ savings have been forecast to be available across the estate in OOH consumption alone. The added benefit of these alerts is to maintain the efficiencies already discovered. As well as finding further areas to save costs, it will also ensure savings are maintained.

Whilst we currently operate at a set threshold, this threshold could be considered lenient. A future aim will be to reduce the benchmark and achieve a higher level of efficiency. As well as reducing the sites OOH threshold kWh being used to raise alerts, one area in which to tighten the benchmark is to adjust the filtering rules currently set to streamline the alerts. This will mean targeting sites sooner after alerts are raised and discover perhaps alternative controls issues.

This project has helped bring into light further projects that could be undertaken such as a CAPEX initiative bringing mezzanine AC units under the control of the BMS during the warmer months, highlighted significantly by our recent warm summer.

[siteorigin_widget class=”SiteOrigin_Widget_Button_Widget”][/siteorigin_widget]

Get In Touch