How does EnergyLabs calculate accurate energy savings?

Berlin. September 2021. With our solution ActiveEnergy, saving energy becomes as easy as making a cup of tea. After our AE PowerStation is installed, it does not require any further actions from our partners and customers. Based on voltage optimization and machine learning, ActiveEnergy unlocks every energy-saving opportunity in a building which reduces the overall carbon footprint. In the following, we will look at how EnergyLabs ensures that the energy savings we report are correct.

Transparency is one of our key values at EnergyLabs. We try to make sure that our partners understand every step of the energy-saving process. Additionally, EnergyLabs ensures that they always have access to their building’s real-time data. To do so, we provide our customer portal, the AE Cloud. Part of the transparency key value is to further elaborate on our calculation process of the achieved energy savings.


Is the energy savings calculation an industry standard?

We have adapted an industry-standard procedure, the “A/B Test”. Thus, we make sure that we have correct reporting on the savings. In A/B-Testing, two different processes under the same conditions are directly compared. The ActiveEnergy solution always includes the measurement of the energy consumption in a building. While EnergyLabs performs the A/B Tests, the AE PowerStation is periodically turned off and on again, while the measuring continues. The AE PowerStation is our hardware that is necessary for the voltage optimization process. The A/B Test and the direct comparison allow us to measure the exact impact of the AE PowerStation on the energy consumption of the building.

MobileHub and toolbox inside in front of wall

The Energy Savings Calculation Process


Imagine a Monday morning. The AE PowerStation has been working overnight, adapting the voltage in the building constantly to an optimal level. At 10 AM, the AE PowerStation is disconnected for 10 minutes in total. We call that “switching into bypass mode„. The building’s power consumption shows to be 100 Kilowatt (kW) in this 10-minute-bypass-mode.

After reactivating the AE PowerStation, the energy consumption decreases by 8 kW, to a level of 92 kW. EnergyLabs can now list this value as an energy saving of 8%. It is necessary to repeat this process between 15-20 times a month, using different time intervals of duration (from 1 to 20 minutes). The energy savings over these time periods are then aggregated. EnergyLabs will regularly report the values to the partners in specific energy savings reports.  

How does the A/B Test work?

So, let’s look a little further into the energy savings calculation process. To measure the correct energy savings, it is necessary to measure the energy consumption twice. We measure once with the AE PowerStation running and once without the AE PowerStation running. When the AE PowerStation is not running, it is in bypass mode. Afterwards, the measurements and energy consumptions in kilowatt (kW) are compared. It is important to note that these measurements must take place at times that usually have a similar or the same energy consumption.

EnergyLabs has adapted the process to the demands of the ActiveEnergy solution. This minimizes the downtime of the AE PowerStation and the loss of savings. At the same time, it maximizes the accuracy and transparency for our customers.

How do you identify matching time slots?

As mentioned before, two time slots of the same length and ideally, the same energy consumption, are needed. To find these matching time slots, EnergyLabs analyses the energy consumption over time, based on machine learning.

The following example in Figure 1 shows two repetitive and consecutive time slots. This also means that the power consumption is constant over a longer period.  As a result, tests A and B can be executed right after each other. The time slots occur daily and can be used for the EnergyLabs A/B test.

Consecutive time slots to calculate energy savings


Figure 1: consecutive time slots


Figure 2 shows an example of non-consecutive time slots. In this case, the slots chosen for comparison do not occur one after the other and are usually separated by a period of time.

Non-consecutive time slots to calculate energy savings

 Figure 2: non-consecutive time slots


AE PowerStation: active vs. bypass Mode

After defining the time slots, as mentioned before, EnergyLabs tracks both the consumption with and without voltage optimization. To do so, we set the AE PowerStation into bypass mode for the first defined point in time of a matching pair. During the second defined time slot, the AE PowerStation is active and running, meaning the hardware performs voltage optimization as usual.

Figure 3 shows the comparison in the measured energy consumption for the previous four example pairs. The red color indicates that the AE PowerStation is in bypass mode during the first measurement. The green color shows that the hardware is active during the second measurement and performs voltage optimization. Energy Savings for chosen time slots

 Figure 3: energy savings for chosen time slots


How often does EnergyLabs perform the AB-Test?

The energy savings calculation process is continuously repeated but uses different time intervals and varies in duration (1-20 minutes). The process takes place 15 to 20 times per month. The result is an averaged energy savings percentage. The higher the number of time slots for A/B tests, the higher is the probability of the saving rate to be accurate. EnergyLabs‘ focus is to find a good balance between a high probability of our saving rate, while also minimizing the time in which the AE PowerStation is in bypass mode. Having an average saving rate, it is possible to reconstruct the original consumption over time and compare it to the measured consumption with the AE PowerStation being active.

Calculating the exact Energy Saving Rate


Following, Figure 4 and Figure 5 show the energy consumptions with and without the AE PowerStation running, still referring to the previously discussed examples. As you can see, there is a difference in the consumption in kilowatts when the AE PowerStation is active (green) compared to it being in bypass mode (red).


Energy consumption with and without AE PowerStation (Example 1)

Figure 4: Energy consumption with and without AE PowerStation (Example 1)Energy consumption with and without AE PowerStation (Example 2)Figure 5: Energy consumption with and without AE PowerStation (Example 2)


Based on this comparison, EnergyLabs calculates the energy-saving rate for the customers, ensuring very high accuracy and transparency.

Energy Savings Value Calculation

Using the exact saving rate and the accumulated monthly energy consumption, EnergyLabs calculates the energy savings per month. In the next step, the specific prices that our customers pay for their energy can be considered. Calculating their monetary savings allows individual and specific saving predictions for the next five years.


Pictures and Figures: EnergyLabs and Unsplash