Algorithms¶

This tab summarizes the value offered by the analysis function, the component types the analysis is recommended for, and the conditions checked by the analysis.

Expect a short case-study on how the analysis function was applied during development or a test bench.

Results of analytics functions are structured to deliver simple-to-navigate insights and fast-to-apply measures on how to improve operational performance.

Therefore, each result regardless of the analytics function includes

• One qualitative warning level, aka. signal color,
• One interpretation,
• Zero to one summary KPI,
• Zero to n recommendations,
• Zero to n KPIs,
• Zero to n plots, and
• Zero to n timeseries.

These categories are explained below. While the warning level, interpretation, and recommendation are specified for all analysis functions equally, summary KPIs, KPIs, plots, and time series differ between each analysis function.

Warning level

The warning level represents the urgency of looking into the analyzed condition and taking action to improve it. It can have one of these traffic light states, but not every analysis makes use of the full spectrum:

Interpretation

The interpretation delivers a summary of the observed performance and state of the condition analyzed. It describes either a symptom of a suboptimal operation or a condition that could be identified.

Recommendation

Recommendations are summarized in a list of zero (for sufficient operational performance) to n measurements on how to improve the operational performance. The recommendations either help by providing information on how to correct the source of the symptom itself or on how to narrow down to its root cause.

Plots

It is easy to understand complex data insights with the right visualization. Thus, many .analytics algorithms offer plotting recommendations including the type of plot, data, and labeling. This structured and well defined plotting data makes it easy to transport data insights.

Available plot types

Included data structurs

Summary KPIs, KPIs, and time series

KPIs and time series offer insights and transparency, while the most important KPI is offered as summary KPI including its current value and KPI development comparred to the previous analysis. They enable reporting and manual investigation of the operational behavior of the component or system analyzed. KPIs and time series are highly individual for each function and are explained in the respective specification of each analysis function in Results.

The Components tab contains the information on how to configure the algorithm for the specific components.

• The components the analysis function is available for,
• The pins of the components which need to be mapped, and
• The attributes of the component required.

The Application tab provides information on the application of the analysis function.

• Recommended time span: Most of the analysis functions have a sweet spot for the amount of historical data required to derive accurate results.
• Recommended repetition: Components of building energy systems are subject to seasonal effects and wear out. Follow the recommended repetition to ensure the analysis is performed as often as necessary, without risking blind spots during the continuous monitoring of the system.

Value

• Avoids alarm messages being overlooked
• Identifies faulty components
• Can reduce component wear-and-tear
• Can increase energy efficiency

Recommended for components

Any component or subsystem which could have an alarm or error message such as:

• Fans
• Heat pumps
• Thermal control loops

Checked conditions

• Last state of alarm message
• Relative duration of alarm message
• Total duration of alarm messages
• Total occurrences of alarm message

The Alarm State analysis was performed on a component for February 2020. The error message is active at the beginning of the time period and then about twice a week after that.

Figure 1: Component error message for the month of February 2020

The analysis returns a red warning message to indicate that the error message over the time period is suboptimal. This is because the error message is active for a significant percentage of the total time.

Signal colors

Interpretations

Recommendations

KPIs

Summary of alarm messages

Air cooler

Air heater

Boiler

Combined heat and power

Compression chiller

Cooling circuit

Fan

Heat pump

Heat transfer unit

Heating circuit

Recooling plant

Thermal control loop

Recommended Time Span

1 month

Recommended Repetition

Every month

• After changes of operational modes, e.g., transfers to heating mode
• After changes in the control system
• After maintenance of replacements

Value

• ESG-Reporting and monitoring of CO₂ emissions
• Insights in CO₂ performance of the building
• CO₂ emission benchmarking
• Trace back CO₂ emissions to energy sources

Recommended for components

Complete buildings or energy systems

Checked conditions

• CO₂ emissions of the building or system
• Benchmarking
• Emission by energy source

This example shows the results of an CO₂ Emission analysis displayed in the Asset Overview.

Figure 1: CO₂ emissions Asset Overview result overview

Above all cards the interpretation text of the result is highlighted with the signal color of the result. In the top left the current KPI and the monthly trend is shown. On the right side of that you will find the recommendations of this result as well as further information about the function.

The three plots below are:

1. The building benchmark

Figure 2: CO₂ emissions Asset Overview plot bullet graph

The current KPI is plotted inside a bullet graph between the limits of the signal color.

2. Compared to last months

Figure 3: CO₂ emissions Asset Overview plot month comparison

The plot shows the current KPI of the last three months, including the month of the analysis.

3. By energy source

Figure 4: CO₂ emissions Asset Overview plot energy source

The plot shows the distribution of CO₂ emissions by energy source.

Signal colors

Interpretations

Recommendations

KPIs

Summary KPI

Timeseries

The timeseries is saved as a virtual datapoint. Every run of the analysis will add the summary KPI (current value) as an observation at the start of the analysis period to the timeseries.

Plots

Building

Sub-components

At leat one sub-component is needed to get a result for this analysis function. Only the necessary meter components need to be added to the project.

Cold meter

Electricity meter

Gas meter

Heat meter

Recommend Time Span

1 month

Recommended Repetition

Monthly

• Continuously monitor CO₂ emmisions
• Track success of a decarbonization measure

Value

• Increase lifetime of valve, dampers, and adjacent components
• Avoid spontaneous failures
• Reduce energy consumption
• Reduce noise pollution

Recommended for components

Any liquid media supply system, such as:

• Thermal control loop with 2-way valve and pump

Checked conditions

• Process value of the control loop is oscillating
• Process value of the control loop is not or to a negligible degree oscillating
• Condition checks on times of components operation

For this example, we analyzed the temperature control loop of a supply air volume flow, which provides fresh air and heating to a large salesroom. Figure 1 shows a plot of the process value of the control loop, the outlet temperature. The plot shows an oscillation of the outlet temperature during periods of operation.

Figure 1: Oscillating processes value during operation

Figure 2 is a zoom of figure 1 to analyze the oscillation in more detail. The trajectory of the process value is common for control loops oscillating at medium frequency.

Figure 2: Oscillating process value during operation in detail

The Control Loop analysis evaluated this oscillation as significant and assigned it the signal color yellow. Recommendations are made on how to adjust controller parameters for a smoother operation.

Signal colors

Interpretations

Recommendations

Air handling unit

Cooling circuit

Heat transfer unit

Heating circuit

Thermal control loop

Recommend Time Span

1 day to 1 week

Recommended Repetition

Weekly

• After changes of operational modes, e.g., transfers to heating mode
• After changes in the control system
• After maintenance or replacements

Value

• Better understanding of the system's operating modes
• Detecting of faulty operation modes
• Lower the energy costs of your system through the optimization of the detected operating modes

Recommended for components

• Air handling unit

Checked faulty conditions

• Air heating without full use of the heat recovery system
• Simultaneous heating and cooling

The sequence analysis was applied to an air handling unit with an analysis period of one week. We can see that the unit has been operating in the faulty air heater mode for a very long time. This suggests that the heat recovery is not fully utilized before the air heater is switched on.

Connected sub-components:

• Supply fan
• Air heater
• Heat recovery system

Figure 1: Operating state of sub-components

Signal colors

Interpretations

Recommendations

KPIs

Faulty operating states

Normal operating states

Air handling unit

Sub-component supply fan

Sub-component exhaust fan

Sub-component air cooler

Sub-component air heater

Sub-component air re-heater

Sub-component heat recovery system

Sub-component supply humidifier

Sub-component exhaust humidifer

Sub-component recirculation damper

Recommend Time Span

1 day to 1 week

• Analysis is useful for all days with system operation

Recommended Repetition

Every month

• After changes of operational modes, e.g., transfers to heating mode
• After changes in the control system of the ventilation systems
• After maintenance or replacements in ventilation systems

Value

• Avoids damage to rooms due to condensation

Recommended for components

• Rooms

Checked conditions

• Duration of dew point alert signal
• Duration in which the room temperature is between 2 °C and 4 °C above the dew point temperature
• Duration in which the room temperature is within 2 °C of the dew point temperature

The Dew Point Alert analysis was performed on a room for a week in February 2020. For this particular room, a dew point alert message is available but no temperature and relative humidity data. As is shown in figure 1, the dew point alert signal is only active for a very short amount of time during the week.

Figure 1: Dew point alert for one week in February 2020

The analysis returns a red warning message to indicate that the dew point alert was active during a portion of the time period. This suggests that the condensation may have formed in the room. Note that only "dew point alert" KPIs are generated since no temperature and humidity data are available in this example.

Signal colors

Interpretations

Recommendations

KPIs

The KPIs which are generated by this analysis depend on the information available in the analysis. The "dew point alert message" KPIs are generated if a dew point alert message is available. The condensation risk KPIs are generated using room temperature and relative humidity.

The condensation risk is evaluated as moderate if the room temperature is between 2 K and 4 K above the dew point temperature. A high condensation risk is when the room temperature is within 2 K of the dew point temperature.

Dew point alert

Room

Recommend Time Span

1 week to 1 month

Recommended Repetition

Every month

• After changes of operational modes, e.g., transfers to heating mode
• After changes in the control system
• After maintenance and replacement

Value

• Detect AHU fans that are not controlled
• Reduce costs through better fan speed control

Recommended for components

• Air handling unit
• Fan

Checked conditions

• Stationary fan speed

In this example, we look at a Dynamic Control analysis of the historic 7-day fan speed. While the operating message (grey in the plot below) shows the times when the fan was operated, the fan speed (blue in the plot below) corresponds to the speed or load setting of the fan.

Figure 1: Speed and operating message of a 7-day fan speed analysis

From the analysis results, we can see that the fan was operated for 6 hours out of the 168 hours of the week or 3.57 % of the week. Additionally, we get statistics of the fan speed, e.g., the fan was operated at an average of 40 % load.

This corresponds to a static fan speed setting that is currently not controlled. To improve energy efficiency and thermal comfort you can consider different control strategies outlined in the recommendations.

KPI - Statistics

Signal colors

Interpretations

Recommendations

KPIs

Statistics

Air handling unit

Fan

Recommended Time Span

1 week

Recommended Repetition

Every month

Value

• Insights into the operating times of elevators over weekdays
• Insight into the total availability of elevators

Recommended for components

Buildings with elevators

Checked conditions for operating time calculation

Checked conditions for availability calculation

An analysis on a building with elevators is carried out for an analysis period of 3 weeks.

KPI - Statistics

Bar chart Average Elevators Operating Time

Signal colors

Interpretations

Recommendations

KPIs

Summary KPI

Plots

Building

Sub-components

At leat one sub-component is needed to get a result for this analysis function.

Elevator

Recommend Time Span

1 week

Recommended Repetition

weekly

• Continuously monitor availability of elevators
• Track operating times of your elevators

Value

• ESG-Reporting and monitoring of energy consumption
• Insights into the energy consumption of the building
• Energy consumption benchmarking
• Trace back consumption to the energy source

Recommended for components

Complete buildings or energy systems

Checked conditions

• Energy consumption of the building or system
• Benchmarking
• Energy consumption by energy source

This example shows the results of an Energy Consumption analysis displayed in the Asset Overview.

Figure 1: Energy Consumption Asset Overview result overview

Above all cards the interpretation text of the result is highlighted with the signal color of the result. In the top left the current KPI and the monthly trend is shown. On the right side of that you will find the recommendations of this result as well as further information about the function.

The three plots below are:

1. The building benchmark

Figure 2: Energy Consumption Asset Overview plot bullet graph

The current energy consumption is plotted inside a bullet graph between the limits of the signal color.

2. Compared to last months

Figure 3: Energy Consumption Asset Overview plot month comparison

The plot shows the energy consumption of the last three months, including the month of the analysis.

3. By energy source

Figure 4: Energy Consumption Asset Overview plot energy source

The plot shows the distribution of energy consumption by energy source.

Signal colors

Interpretations

Recommendations

KPIs

Summary KPI

Timeseries

The timeseries is saved as a virtual datapoint. Every run of the analysis will add the summary KPI (current value) as an observation at the start of the analysis period to the timeseries.

Plots

Building

Sub-components

At leat one sub-component is needed to get a result for this analysis function. Only the necessary meter components need to be added to the project.

Cold meter

Electricity meter

Gas meter

Heat meter

Recommend Time Span

1 month

Recommended Repetition

Monthly

• Continuously monitor energy costs
• Track success of a energy savings measure

Value

• ESG-Reporting and monitoring of energy costs
• Insights into the energy costs of the building
• Energy costs benchmarking
• Trace back costs to the energy source

Recommended for components

Complete buildings or energy systems

Checked conditions

• Energy costs of the building or system
• Benchmarking
• Energy costs by energy source

This example shows the results of an Energy Costs analysis displayed in the Asset Overview.

Figure 1: Energy Cost Asset Overview result overview

Above all cards the interpretation text of the result is highlighted with the signal color of the result. In the top left the current KPI and the monthly trend is shown. On the right side of that you will find the recommendations of this result as well as further information about the function.

The three plots below are:

1. The building benchmark

Figure 2: Energy Cost Asset Overview plot bullet graph

The current energy cost is plotted inside a bullet graph between the limits of the signal color.

2. Compared to last months

Figure 3: Energy Cost Asset Overview plot month comparison

The plot shows the energy costs of the last three months, including the month of the analysis.

3. By energy source

Figure 4: Energy Cost Asset Overview plot energy source

The plot shows the distribution of energy costs by energy source.

Signal colors

Interpretations

Recommendations

KPIs

Summary KPI

Timeseries

The timeseries is saved as a virtual datapoint. Every run of the analysis will add the summary KPI (current value) as an observation at the start of the analysis period to the timeseries.

Plots

Building

Sub-components

At leat one sub-component is needed to get a result for this analysis function. Only the necessary meter components need to be added to the project.

Cold meter

Electricity meter

Gas meter

Heat meter

Recommend Time Span

1 month

Recommended Repetition

Monthly

• Continuously monitor energy costs
• Track success of a energy savings measure

Value

• Reduced operational costs
• Detection of broken components

Recommended for components

• Combined heat and power

Checked conditions

• Efficiency of component
• Power to heat ratio (for component combined heat and power)

The Energy Efficiency analysis was applied to a real combined heat and power plant and the pins operating message, generator power, heat flow, and fuel power were mapped. Figure 1 shows the time series recorded for an exemplary period of 7 days. The grey shaded periods correspond to the operation of the plant.

Figure 1: Period of one week for which the energy conversion is not acceptable

A very low generator power compared to the heat flow and the fuel power indicates a failure of the generator component. The automated interpretation confirms our visual analysis of the time series shown in the figure, summed up by the qualitative warning level “red”. It also provides recommendations to address the problem.

Signal colors

Interpretations

Recommendations

KPIs

Energy Performance KPIs

Combined heat and power

Recommended Time Span

1 day to 1 week

Recommended Repetition

Every month

Value

• Ensures filter is serviced when required
• Improves energy efficiency

Recommended for components

• Filter

Checked conditions

• Filter contamination
• Expected time till filter service or replacement

In this example, the filter contamination of an exhaust air filter of an air handling unit was analyzed over a period of four months. As can be seen in figure 1, filter contamination gradually increases over the analyzed period.

Figure 1: Filter contamination over a four month period

The signal analysis returns a green signal color since there is a significant amount of time before the filter is fully contaminated.

Signal colors

Interpretations

Recommendations

KPIs

Filter

Recommended Time Span

1 month to 6 months

Recommended Repetition

Twice a month

Value

• Linear model representing the heat demand of the building
• Determination of a basic heat load e.g. for hot water energy demand
• Determination of a heating limit temperature
• Determination of a maximum heating power within one period
• Determination of zero heating power duration within one period

Recommended for components

Building with gas heating or district heating

Checked conditions

• Heat energy consumption of the building
• Comparison of heat energy consumption between different periods
• Energy saving calculation of reference period to analysis period

This example shows the monthly sampled heat energy of the main heat meter over the period of more than one year. As shown in Figure 1, a heating period between November and April and a period without required heating were selected.

Figure 1: Heat meter energy samples with monthly resolution

Figure 2: Annual duration curve of the heat power

[]
Figure 3: Heat demand curve with base load power

Figure 4: Heat demand result

Signal colors

Interpretations

Recommendations

KPIs

Summary KPI

Building

Sub-components

Weather Station

Adding the weather station is mandatory. The analysis uses the outside air temperature as one of two input values.
The larger the temperature interval, the more accurate the heat demand analysis. If there are several weather stations, the first one is selected in alphabetical order of the component in project name.

At least only one meter component is needed to get a result for this analysis function.

Gas meter

Heat meter

Recommend Time Span

1 year, but at least one full heating period and one period without heating demand

Recommended Repetition

monthly

• Continuously monitor energy savings in comparison to the reference period

Value

• Hourly operating point simulation depending on heat demand and environmental heat
• Choice of different heat pump models and technologies from several manufacturers
• Energy saving calculation and conversion in cost and CO₂ emission savings potential
• Investment calculation depending on
  • cost savings
  • interest rate
  • depreciation rate (optional)

Recommended for components

Building with gas heating or district heating

Checked conditions

• Heat energy consumption of the building
• Efficiency potential of heat pump usage
• Payback period for heat pump investment

This example shows a simulation of an air/water heat pump. For a brief technical explanation, see Figure 1.

Figure 1: Heat pump schematic

Figure 2: COP map air/water heat pump

Figure 3: Test reference year outside air temperature

Figure 4: Heating power (TRY2015)

Figure 5: Supply temperature and sink temperature (TRY2015)

Figure 6: Simulation results at different supply temperature and nominal power

Technical heat pump KPIs

While the maximum demanded heat power is 530 kW, the optimum heat pump power is selected to 360 kW at a design supply temperature of 50 °C. Thus, the majority of all operating points in the test reference year can be covered with 1400 MWh, especially for this heating system. The remaining heat energy about 100 MWh is provided by a heating element with 170 kW. With an annual cop about 2.53, the electric energy consumption is about 570 MWh.

Economic heat pump KPIs

Assuming an electricity price and a heat energy/gas price of the existing heating system, a cost saving can be calculated. In addition to the investment cost of the heat pump installation, other investment parameters can be taken into account. This includes:

-   Price increase rate of energy costs
-   Calculation interest rate
-   Depreciation period

With an installation price of 800 €/kW the heat pump, including installation and material, the total investment is about 295700 €. Assuming an electricity price about 0.24 €/kWh and a heat energy price about 0.12 €/kWh, we get a total cash flow about 298600 € after 5 years. The cash flow considers a yearly price increase rate of energy costs about 3% and a calculation interest rate about 12%. Thus, the annual cash flow is discounted and accumulated to a total cash flow. Where the CO₂ emission factor of the current heating system is assumed with 0.2 kg/kWh, the CO₂ emission factor of the electric net supply is assumed with 0.4 kg/kWh. Nevertheless, a CO2 saving of 63 t/a is achieved, since the primary amount of energy is significantly lower with 570 MWh.
As a result, a payback period of 5 years is sufficiently small to consider a closer look.

The interpretation would be:
The conditions for a heat pump investment are suitable

The recommendation would be:
It might make sense to further review the energy system. For this you could contact a specialist planner.

Signal colors

Interpretations

Recommendations

KPIs

Summary KPI

Building

Sub-components

Weather Station

Adding the weather station is mandatory. The analysis uses the outside air temperature as one of two input values.
If the temperature value of the current weather station does not seem reliable or plausible, it is recommended to create a weather station component Analyse Wetterstation. The pin Temperatur should be mapped with the weather service data point to get a reliable and gap free time series. The weather station is sorted alphabetically. The first one will be chosen for the analysis.
The larger the temperature interval, the more accurate the heat demand analysis.

At least only one meter component is needed to get a result for this analysis function.

Gas meter

Heat meter

Recommend Time Span

1 year, but at least one full heating period and one period without heating demand

Recommended Repetition

monthly

• Continuously monitor the energy consumption for a replacement investment recommendation

Value

• Detect operating conditions of AHUs that are not appropriate to the outside air conditions
• Avoids unnecessary changes in humidity, which cost a lot of energy
• Verifies sufficient supply air humidity

Recommended for components

• Air handling units with humidity conditioning

Checked conditions

• Compare actual operating hours with humidification, dehumidification, and no operation with the corresponding expected hours

This example shows a week of analysis for a summer scenario in July. The AHU is operating throughout the week. Relative humidity conditions are displayed in red and orange, green and blue are temperature conditions and brown and purple are the water load conditions.

Figure 1: Analysis of humidity, temperature and water load conditions over a whole week

The analysis uses two positions, intake (outside conditions) and outlet (supply conditions) to calculate water loads. A difference in these water loads corresponds to the pink line at the bottom. The operating hours will now be divided into three categories. Hours of humidification, hours of dehumidification, and hours of neither humidification nor dehumidification. These values are then compared to the expected hours in these categories derived from outside conditions. The total hours of correct operation (according to the expectation) are then evaluated for a recommendation.

Signal colors

Interpretations

Recommendations

KPIs

Operating Conditions

Humidity conditioner

Recommended Time Span

1 week

Recommended Repetition

Every month

• After changes of operational modes
• After changes in the control system

Value

• Lower operating costs
• Higher energy efficiency
• Peak energy consumption reduction
• Longer equipment and component lifetimes
• Smoother system integration

Recommended for components

Energy conversion plants and components with high start-up energy consumption or wear, such as

• Heat pump
• Combined heat and power
• Boiler
• Fan

Checked conditions

• Short cycling of component operation, evaluated component-specific
• Long cycling of component operation, evaluated component-specific
• Expected cycling of component operation, evaluated component-specific
• Condition checks on times of components operation

The Operating Cycle analysis was applied to a heat pump. Thus, a heat pump component model was instanced and the respective datapoint mapped to the pin operating message. Figure 1 shows the time series recorded for an exemplary period of 1 week in winter.

Figure 1: Operating message and cycle behavior of heat pump

Short shut-down times are observed between periods of duty indicating excessive start and stop processes of the heat pump. This not only leads to energy losses and electricity consumption peaks, but also increased wear-and-tear of the heat pump's compressor.

The attributes "Coefficient of performance", "Nominal heat production" and "Electricity price" are set to 4, 1000kW, and 0.18€/kWh. This enables Economic KPIs with the accuracy level "High". The calculations show daily start-up costs of 66 €/d and costs of one start to be 2.75 € (start-up cost).

The automated interpretation confirms our visual analysis of the time series shown in the figure, summed up by the qualitative warning level "yellow". The recommendations provide further instruction on how to isolate and fix the cause for the increased number of start and stop processes. Furthermore, the result offers an advanced set of KPIs, providing additional insights into the cycle behavior of the heat pump.

Signal colors

Interpretations

Recommendations

KPIs

Operating Time and Operating Cycles

Economic KPIs

Economic KPIs estimate the economic optimization potentials of the observed operational state. The KPIs indicate the total startup costs based on KPI "starts" and the costs of one cycle of the investigated component.

The Economic KPIs are provided for the components:

• Boiler
• Combined heat and power
• Heat Pump

Their availability and accuracy depend on the component's mapping. The analysis function always determines the highest possible accuracy.

Accuracy Levels:

The following tables summarize the pins and attributes required to achieve various accuracy levels for the economic KPIs.

High

Medium

If the available attributes and mapped pins are not sufficient enough to reach a "Medium" accuracy for the economic KPIs, no economic KPIs are calculated.

Air handling unit

Boiler

Compression chiller

Combined heat and power

Fan

Heat pump

Thermal control loop

Recommend Time Span

1 day to 1 week

Recommended Repetition

Every month

• Cycle rates have a strong seasonal effect
• Frequent repetition allows to identify operational bad points
• After changes of operational modes, e.g., transfers to heating mode
• After changes in the control system
• After maintenance or replacements

Value

• Lower operating costs
• Lower energy consumption

Recommended for components

Heat and cold distribution systems, energy conversion plants, and indoor areas, such as

• Thermal control loops
• Boilers
• Rooms

Checked conditions

• Existence of a load reduction period, e.g., night-time temperature reduction for heating
• Condition checks on times of components operation
• Estimation of times when the load can be reduced according to a user-defined schedules

This example shows the results of a Reduced Load analysis performed on a heating circuit.

Figure 1: Temperature setpoint of the heating circuit

Schedule

Signal colors

Interpretations

Recommendations

KPIs

Identification of reduced load mode

Statistics of temperature level shift

Schedule operating times

KPIs of this category analyse if the load reduction is in accordance to a schedule and if there are further savings by adjusting/implementing a load reduction schedule.

Boiler

Combined heat and power

Heating circuit

Room

Thermal control loop

Recommend Time Span

1 day to 1 week

Recommended Repetition

Every 3 months

• After changes of operational modes, e.g., transfers to heating mode
• After changes in the control system
• After maintenance or replacements

Value

• Higher occupant comfort, health and performance

Recommended for components

• Room

Checked conditions

• Identification of higher room ventilation needs
• Detection of poor air quality during peak usage times
• Condition checks on times of components operation
• Sensor callibration check and automated correction

In order to calculate the air quality index, the measured CO₂ (PPM) values are first sorted in a descending order. For an example timeseries, this can be seen as the grey dashed line in Figure 1.
The time period of the recorded data is then normalised to a percentage of time (see x-axis of figure 1) and the timeseries is corrected (blue line in figure 1). This is done by calculating a correction factor at the 98.5 percentile (pink bar), so that CO₂ at this point is 450 ppm. The 450 ppm value is choosen to reflect outside conditions and it is expected that the 98.5 percentile PPM value should not be below outside conditions. The entire timeseries is raised or lowered by this factor (see blue vs. grey line in Figure 1). The value of this correction offset factor is returned as a KPI "Correction offset CO2 sensor".
Following this, the air quality index is now calculated with the CO₂ value at the upper 1.5 percentile of the corrected CO₂ timeseries (yellow bar).

Figure 1: Measured, corrected and ordered CO₂ timeseries and relevant percentiles for the air quality calculation

Using this CO₂ value and the curve shown in figure 2, an overall air quality index between 0-100% is calculated.
For a standard week of 168h, the absolute time for the 1.5% percentile is approximately 2.52h. The use of the 1.5% percentile value ensures that a warning is triggered if the room has a high CO2 reading longer than a cumulative period of 2.5h. However, no warning is triggered if extremely high values occur for only a very short time, as this could be due to a possible measurement error.
The default value of the "1.5% percentile" can be changed via the attribute "Percentile value for calculating air quality". There are also 4 attributes to adjust the two points 1 and 2 of the curve in Figure 2 as well as the limits for the red and yellow signal colors.
For futher default values see attributes under the section components.

Figure 2: Curve for calculating the overall air quality percentage (blue) and limits for signal colors and classifications of air quality

Signal colors

Interpretations

Recommendations

KPIs

Operational KPIs

Default values KPIs

Returns some relevant default values that were used to calculate the air quality

Air quality KPIs

Statistical KPIs of CO₂ concentration

Providing deeper insights into the carbon dioxide concentrations over the analyzed period.

Room

Recommend Time Span

1 day to 1 week

• Utilize on days with room occupation

Recommended Repetition

Every month

• After changes of room occupation or usage
• After changes of operational modes, e.g., transfers to heating mode
• After changes in the control system of the ventilation systems
• After maintenance or replacements in ventilation systems

Value

• Lower operating times of HVAC components
• Lower energy consumption
• Lower maintenance costs due to less component operating time

Recommended for components

Any HVAC component or room whose usage follows a recurrent schedule, such as:

• Air handling units
• Thermal control loops
• Office rooms
• Sales rooms

Checked conditions

• Component operation outside a user-defined schedule
• Component operation during a user-defined schedule
• Condition checks on times of components operation

This example shows a Schedule analysis for a component "fan" connected to a supply fan operating message of an HVAC machine. The switch on/off times of the machine are shown as a blue line in figure 1, blue regions in the background correspond to the expected schedule.

Figure 1: Operating times of component and reference schedule

A reduction of ~9% of the total operating time is possible, as can be seen in the table of KPIs below. With the help of the plot we can also see, that the times where we can reduce the operating time are distributed over the workdays of the week. Furthermore, the mapping of the component enables the determination of economic KPIs with a "low" accuracy level. Assuming a nominal power consumption of 3 kW and the electricity price to be 0.18 €/kWh, the energy costs outside the scheduled operating times are calculated. This results in weekly energy costs of 3.75 € which fall outside of the schedule.

Signal colors

Interpretations

Recommendations

KPIs

Schedule KPIs

Operating KPIs

If there are multiple sub-components with operating times a prefix is added to the kpi identifer. For example the AHU will generate operating times for Supply fan and Exhaust fan.

Energy KPIs

If there are multiple sub-components with operating times a prefix is added to the kpi identifer. For example the AHU will generate energy consumption KPIs for Supply fan and Exhaust fan.

Environmental KPIs

Economic KPIs

Savings potential KPIs

If there are multiple sub-components with operating times a prefix is added to the kpi identifer. For example the AHU will generate energy consumption KPIs for Supply fan and Exhaust fan.

These KPIs are used to show the savings potential of the component in the analysis period. The deciding input is how many operating hours could have been reduced adhering to the schedule thus savings energy, CO2 emissions and costs.

Yearly savings potential KPIs

These KPIs estimate the savings potential extrapolated for a whole reference year. The reducible hours calculated from the analysis period are used to extrapolate the whole year.

The Economic KPIs, Savings potential KPIs and Yearly savings potential KPIs are provided for the components:

• Air handling unit

Options for savings calculation

The following tables summarize the possible pin and attribute combinations that are needed to calculate the yearly savings potential.

Configuration for general schedule analysis

Configuration heat recovery system

Configuration for savings calculation for heat

Configuration of savings calculation for cold

Savings for electrical energy

Savings for thermal energy - heat

Savings for thermal energy - cold

Air handling unit

Boiler

Combined heat and power

Fan

Heat pump

Thermal control loop

Room

Recommend Time Span

1 week

Recommended Repetition

Every week

• After adjustment of usage times of the analyzed component
• After changes of operational modes, e.g., transfers to heating mode
• After changes in the control system
• After maintenance or replacements

Value

• Confirm normal operation of sensors
• Identify faulty measurement setups inside your building automation system
• Detection of permanently manual overwritten sensors causing permanent manipulation of control loop

Recommended for components

Any component with sensors measuring physical quantities.

Checked conditions

• Measurements of a sensor lie within a reasonable range
• Detects constant observation for sensors which expect volatile trajectories

For this example we are looking at a temperature sensor for the room air temperature, that is connected to a component "room". The KPIs are generated according to the mapped pins. For this setup we mapped a datapoint to pin "temperature", thus the result contains the three KPIs listed below.

The room temperature is measured by the sensor with values above the plausibility limit of 40 °C. The KPI "pin.temperature.above high limit = 1" indicates that the measured values do not lie within a reasonable range for room temperatures.

If any of the KPIs have the boolean value of 1, a faulty sensor is detected and the signal color red is returned to alarm. A detected fault can be caused by various reasons ranging from manually overwritten sensors over a faulty sensor to a wrong configured measurement system.

Figure 1: Room air temperature over a two day period

Signal colors

Interpretations

Recommendations

KPIs

Sensor Fault