By Kurnia · Published November 13, 2026 · Version 1 · 19 min read
Building energy dashboards
meters and submeters, energy use intensity, baselines and weather normalization, reading waste from trends, and verifying savings
The October electricity bill of a 220-room hotel is 9 percent higher than September’s, exactly one month after a project fitting variable speed drives to the pumps and cooling towers was completed, and the general manager suspects the project failed. The building energy dashboard shows two other causes: October was hotter, and since October 12 the night load has risen by 60 kW because two kitchen exhaust fans run continuously. This chapter covers meters and submeters and the metering hierarchy according to the FEMP metering guide, interval data and what goes on a dashboard, energy use intensity per floor area and per unit of service, baselines and weather normalization with cooling degree days, reading base load, night load, weekends, and step changes from trends, and verifying savings with the savings equation and the conditions for the measurement options from the FEMP M&V guidelines. The official Indonesian energy use intensity criteria could not be checked as of November 2026, so the examples use practice numbers. Installing meters in live panels and changing BMS programs are set out as work for authorized persons.
In early November, the general manager of a 220-room hotel receives an October electricity bill that is 9 percent higher than September’s. A month earlier, the hotel had completed a project fitting variable speed drives to the chilled water pumps and the cooling tower fans, with the promise that electricity use would fall. The question comes at once: has the project failed? The chief engineer opens the building energy dashboard. The monthly chart shows that October was much hotter than September. The 15-minute chart shows something else: since the night of October 12, the electrical load between 23:00 and 06:00 has risen from about 310 kW to 370 kW and has never come down again. The kitchen panel submeter explains why: after the kitchen panel was renovated, the selector switches of two kitchen exhaust fans were left in the hand position, so the fans ran day and night.
That higher bill contains three things at once: weather, new waste, and project savings. The bill alone cannot separate them. Well-organized meters, interval data, a baseline that accounts for weather, and a way of reading trends can separate them, and that is what a useful building energy dashboard contains.
Scope
This chapter covers building energy dashboards from the side of the technician and the person responsible for energy: the arrangement of meters and submeters, interval data, energy use intensity, baselines and weather normalization, reading waste from trends, and checking savings claims. Baseline regression, CUSUM cumulative sums, savings uncertainty, the IPMVP options, and the content of a verification report are covered in full in Chapter 15 · Data & Measurement on measuring and verifying energy savings; this chapter uses its results in buildings. The carbon footprint is in Chapter 19 · Fluids & Thermal, how to present charts and dashboards in Chapter 16 · Data & Measurement, flow measurement and the water meter balance in Chapter 5 · Fluids & Thermal, operating KPIs in Chapter 20 · Laundry, laundry energy efficiency in Chapter 11 · Utilities, and BMS trends and schedules in Chapter 33 on BMS alarms, trends, and schedules. Data points and the points list are in Chapter 32 on BMS architecture.
The official energy use intensity criteria for buildings in Indonesia could not be checked from official documents as of November 2026, so this chapter does not give Indonesian benchmark figures. The numbers in the examples are made up for practice, and US or UK data are identified as data from those countries.
Work outside the technician’s authority: installing or moving meters, current transformers, and pulse counters in live panels, which is work for an authorized electrician with a work permit according to Chapter 76 · Electrical & Control on safe work; changing controller programs, controller schedules, and meter configuration in the BMS, which is the authority of the system integrator; and issuing savings figures for contracts or mandatory reports, which is the authority of the energy manager or energy auditor. The technician reads, records, checks data, and reports.
Meters, submeters, and the metering hierarchy
A utility meter is a meter owned by the energy supplier that is the basis for billing. A submeter is a meter owned by the building that measures part of the consumption behind the utility meter, for example one panel, one system, or one machine. The metering guide from FEMP, the US federal energy management program, arranges metering as a hierarchy: a campus meter for several buildings, a building meter, subsystem meters such as a lighting panel or a chiller plant, branch circuit meters inside a panel, and end-use meters for a single piece of equipment. The lower the level, the higher the cost and the sharper the diagnosis. A building meter with interval data is often enough to “frame the question,” that is, to show which system is behaving oddly, and submeters are then used to answer it.
The hierarchy also provides a check. The sum of the submeters under one main meter is compared with that main meter:
is the energy of the main meter, the energy of submeter , and the energy not measured by submeters, all in kWh for the same period. That residual contains unmetered loads, cable and transformer losses, and meter errors. A residual that is stable from month to month is a property of the building; a residual that jumps points to a faulty meter, a current transformer installed backward, a new load without a meter, or missing data. Chapter 5 · Fluids & Thermal on flow measurement uses the same balance for water meters.
A good meter for a dashboard stores interval data, has a recorded scale factor (for example the current transformer ratio and kWh per pulse), and is connected to the BMS or a meter data system. A mistyped scale factor makes a submeter report ten times too much, and that mistake shows up first as a negative residual.
Interval data and energy dashboards
Interval data are time-stamped consumption values recorded at a fixed interval; modern utility meters generally record every 15 minutes. The FEMP metering guide maps data intervals to their uses.
| Interval | Use according to the FEMP metering guide |
|---|---|
| Annual | compare several buildings and benchmarks, track progress toward goals, set priorities |
| Monthly | compare with monthly weather data, track performance changes |
| Daily | assess heating and cooling, see weekday, weekend, and holiday patterns |
| Hourly | assess the operating schedule from the daily profile, compare occupied and unoccupied hours |
| 15 minutes | as hourly, plus assessing equipment function and specific operating problems |
| Finer | assess equipment function and detect and diagnose faults |
Data accumulates quickly. The same guide calculates that 50 meters with 15-minute data produce 4,800 values per day and more than 1.7 million values per year, so data retrieval and analysis need to be automated. The guide recommends that data be collected and reviewed every day, with simple quality checks: the data was received, its values are within range, and it is complete. Gaps and out-of-range values are alarmed, and data that is substituted or omitted is recorded together with its assumptions. How to clean data is in Chapter 13 · Data & Measurement on cleaning technical data.
An energy dashboard is a display that summarizes meter data into a few charts and figures for daily decisions. A dashboard that is useful to building technicians contains the 15-minute daily profile of the main meter, night load and weekend load, a monthly chart together with weather, monthly intensity, the submeter balance, and an exception list, that is, the meters or days that deviate from normal. A dashboard that shows only today’s figure without a comparison helps no one. Rules for choosing charts and scales are in Chapter 16 · Data & Measurement on presenting technical data.
Energy use intensity
Energy use intensity (IKE, Indonesian intensitas konsumsi energi) is the energy used by a building in a period divided by its floor area:
is the energy over one year (kWh) and the floor area (m²), so IKE is in kWh per m² per year. The FEMP metering guide calls this index a common normalization technique: annual or monthly energy divided by gross floor area. Two things must stay the same every time IKE is compared. First, the definition of area: gross floor area, conditioned area, or lettable area give different figures. Second, the type of energy: an electricity-only IKE cannot be compared with an IKE that adds electricity, LPG, and diesel after converting them with official factors.
IKE compares buildings of different sizes, but does not distinguish buildings of different types or occupancy levels. For hotels, energy per occupied room is often more honest, and for laundries energy per kilogram of linen as in Chapter 11 · Utilities on laundry energy efficiency. The FEMP guide also mentions a benchmark of energy per area per degree day to reduce the effect of climate differences between regions.
The UK uses a similar comparison for public buildings. A Display Energy Certificate (DEC) shows an operational rating from the energy measured by meters over 12 months, compared with a building of the same type with typical performance; a typical building is rated 100, and a building with twice the emissions of a typical building is rated 200. The method allows adjustments for longer occupancy hours, weather and climate, and energy uses that are not typical of the building type. The DEC rating is calculated in carbon dioxide equivalent, which is covered in Chapter 19 · Fluids & Thermal on carbon footprints. The lesson for dashboards is the same: an intensity figure only has meaning when it is compared with a building or period of the same kind and adjusted for clear differences.
Baselines and weather normalization
A baseline is the energy consumption over a reference period, usually 12 months, together with the factors that affect it. In tropical buildings, the largest factor is usually the cooling load, which follows the outdoor air temperature. A cooling degree day (CDD) is the sum of the differences between the daily mean temperature and a base temperature, for each day in a period, whenever the mean is above the base:
is the daily mean outdoor air temperature (°C), the base temperature (°C), and CDD is in degree-days. The supplement to the FEMP M&V guidelines uses a base temperature of 65 °F for CDD in its example model, and suggests a custom base temperature chosen from the behavior of the building itself. For buildings in Indonesia, the base temperature is chosen from the building’s data, as explained in Chapter 15 · Data & Measurement on verifying savings; the examples in this chapter use 24 °C for practice.
The simplest baseline model for a building’s monthly electricity takes the form
with the predicted consumption (kWh per month), the load that does not depend on weather (kWh per month), and the additional energy per degree-day (kWh per CDD). Weather normalization is comparing measured consumption with the consumption predicted by the model for that period’s weather, not with the raw consumption of another month. The FEMP metering guide recommends weather normalization when seasonal comparisons are made.
The model can only be trusted if it fits its data. The supplement to the FEMP M&V guidelines gives example acceptable values for validating regression models: coefficient of determination above 0.75, CV(RMSE) below 15 percent, mean bias error within plus or minus 7 percent, and a t-statistic above 2 for each variable. The same guidance notes that models are generally good only for the range of the variables used to build them. Those figures are suggestions from US government guidance for performance-based contracts, not Indonesian requirements. How to calculate the coefficients, the goodness-of-fit measures, and their uncertainty is in Chapter 15 · Data & Measurement on verifying savings.
Reading waste from trends
The FEMP metering guide divides a building’s daily profile into three parts. Base load is the consumption present 24 hours a day with a similar pattern every day, such as security lighting, ventilation, and plug loads that stay on. Variable load follows occupancy and schedules, such as lighting, computers, and kitchen equipment. Weather-dependent load follows outdoor temperature and humidity and solar gain. The guide suggests routine questions: when does the peak occur, is the base load reasonable, does the profile match operating hours, and is the trend slowly drifting higher.
The case examples in that guide show patterns that can be recognized on any building’s dashboard:
- A daily profile that is flat from day to night shows that the building operates 24 hours or that the night setback function has been disabled; in one building, that function was disabled for night meetings and never restored.
- One night the load does not fall to its usual level, about 21 kW, but stays at 60 to 80 kW; the question raised is which control is in override and what did not switch off.
- After a brief power outage, the daytime and night loads rise and stay up; it turned out that the tenant had installed new equipment.
- On the weekend, the unoccupied daytime load is lower than the night load, so there is about 9 kW that runs only at night and deserves investigation.
- A water meter that records small regular pulses while the building is empty indicates a small continuous leak.
Waste from a load that does not switch off is calculated with
with the increase in power (kW), the hours per day affected, and the number of days, so is in kWh. The FEMP maintenance guide names the most frequent causes: motors, pumps, fans, and AHUs running 24 hours a day when not needed, simultaneous heating and cooling caused by improper setpoints, and overrides that are not released. In tropical hotels, reheat to reduce humidity is genuinely needed in some rooms, so what is looked for is reheat beyond that need.
The dashboard shows when and how much; the cause is found from BMS trends, submeters, and field checks, as in Chapter 33 on BMS alarms, trends, and schedules. The technician reports the finding with its charts. Restoring schedules or automatic functions in the controller is done by an authorized person, and selector switches in panels are returned by an electrician or an assigned operator.
Verifying savings
The FEMP M&V guidelines state that “actual” savings cannot be measured, because savings are the absence of consumption; savings are determined by comparing consumption before and after the measure, with adjustments for changes in conditions. The general equation is:
is the savings, the baseline consumption, the consumption after the measure, and the adjustments for weather, occupancy, and other factors unrelated to equipment performance, all in kWh. In dashboard practice, is the prediction of the baseline model for the conditions of the reporting period, so the weather adjustment is already included in that prediction.
Reading savings from the whole-building meter, called option C, looks simplest because that meter is the basis of the bill. The FEMP M&V guidelines give conditions for option C to be suitable: predicted savings greater than about 10 to 20 percent of the monthly consumption measured by the meter, at least 12 and preferably 24 months of data before the measure, at least 9 and preferably 12 months of performance period data, enough data on the independent variables, and other loads on the meter that are small and constant. The same guidelines give the reasons option C is rarely used: other changes in the building obscure the savings, and building-owned meters need to be maintained and calibrated.
For a measure that affects only one system, such as variable speed drives on chilled water pumps, a submeter on that system gives a sharper answer; this approach is a retrofit isolation as in Chapter 15 · Data & Measurement. For the technician the question is practical: are the expected savings much larger than the monthly variation of the meter being used? If not, the savings will not be visible on the building meter even though they are real.
Worked example
The numbers in the examples below are practice numbers, unless their origin is stated.
Example 1: submeter balance. The hotel main meter records 397,900 kWh in October. Submeters: chiller plant 110,900, AHUs and fan coil units 76,400, lighting 48,600, kitchen 52,300, laundry 41,200, and lifts and pumps 51,500 kWh. The total is kWh, so kWh, that is, percent. If the residual in previous months was 4 to 5 percent, this balance is normal; if it suddenly becomes 12 percent, a meter or a new load needs checking.
Example 2: intensity per area and per occupied room. Last year the hotel used 4,380,000 kWh of electricity with a gross floor area of 23,500 m², so kWh/m² per year. With 220 rooms and 75 percent occupancy, the occupied rooms in a year are , and energy per occupied room is kWh. The year before, 4,510,000 kWh at 70 percent occupancy gives kWh/m² and kWh per occupied room. The area IKE fell by kWh/m², while the energy per occupied room fell more sharply because occupancy rose.
Example 3: monthly cooling degree days. With a base temperature of 24 °C, every day in September had a mean above the base temperature, with a monthly average of 28.0 °C, so . October averaged 29.0 °C over 31 days, so . This shortcut is correct only because no day had a mean below 24 °C; if there is one, CDD is calculated day by day.
Example 4: normalizing the bill for weather. The baseline model from the 12 months before the project is kWh per month. The September prediction is kWh, and the measured value is 365,000 kWh. The October prediction is kWh, and the measured value is 397,900 kWh. The raw increase is percent, but the increase expected from weather alone is kWh out of a measured increase of kWh. October is kWh above the model.
Example 5: a night load that does not drop. The kitchen exhaust fans, which normally stop from 22:00 to 06:00, have run continuously since October 12, adding 60 kW. Over the 20 nights to the end of October, kWh. At a tariff of Rp1,500 per kWh, the cost is rupiah. The night load rose from 310 kW to 370 kW, that is, percent, which is clearly visible in the 15-minute profile even though it is only 2.4 percent of the monthly bill.
Example 6: project savings, building meter versus submeter. Without the fan waste, October would have measured kWh, so the savings according to the building meter are kWh, that is, percent. That figure is far below the threshold of about 10 percent for option C, so it cannot be proven from the building meter. The chiller plant submeter has a baseline of with and CV(RMSE) of 4.8 percent, both meeting the FEMP example validation values. The October prediction is kWh, the measured value is 110,900 kWh, so kWh or percent of chiller plant consumption. The project did not fail; its savings are small relative to the building and were masked by the weather and the kitchen fans.
Stop conditions
- Meter data has to be taken from a meter that is not yet connected, or current transformers would have to be installed or moved in a live panel: stop; that work belongs to an authorized electrician with a work permit.
- The submeter balance jumps, the residual becomes negative, or a meter scale factor is in doubt: do not issue intensity or savings figures; report it to the person responsible for energy and the integrator for checking the meters and their configuration.
- The savings will be used for contract payments, mandatory reports to the government, or public claims: hand it over to the energy manager or a certified energy auditor according to Chapter 15 · Data & Measurement on verifying savings.
- Waste is found in schedules or automatic functions in the controller: do not change the program or schedule yourself; report it with charts to the system integrator or the assigned person.
- A proposed saving comes from switching off ventilation, raising the temperature of rooms whose temperature is controlled for health reasons, or disabling alarms and safety devices: stop that proposal and hand it to the person responsible for the facility and the units concerned.
- The baseline model is used for a month whose weather or occupancy is outside the range of its data, or it does not meet the goodness-of-fit measures: mark the result as an estimate and hand the analysis to the person responsible for energy.
Common misinterpretations
Guess what’s wrong first, then open the explanation.
Common misconception
A higher bill means the energy saving project failed.
Why is it wrong?
The bill mixes weather, occupancy, new waste, and savings. Separating them requires a normalized baseline and submeters.
Common misconception
Our building’s IKE is lower, so our building is more efficient.
Why is it wrong?
IKE is only comparable when the definition of area, the type of energy, the type of building, and the occupancy level are the same.
Common misconception
The sum of the submeters must match the main meter exactly.
Why is it wrong?
There is always a residual from unmetered loads, losses, and errors. What is monitored is the stability of that residual.
Common misconception
A high night load is a property of the building.
Why is it wrong?
A base load does exist, but step increases, odd weekend loads, and flat profiles are signs of waste that can be traced.
Common misconception
Small savings on the building meter mean there are no savings.
Why is it wrong?
Savings of a few percent from one system are lost in the variation of the building meter. A submeter on that system can prove them.
Links to other chapters
Chapter 15 · Data & Measurement on verifying savings provides the statistical methods for baselines, CUSUM, and uncertainty used here. Chapter 33 on BMS alarms, trends, and schedules explains the trends and overrides that are the source of much waste, and Chapter 32 on BMS architecture explains meter points in the BMS. Chapter 16 · Data & Measurement on presenting data helps build the dashboard, Chapter 13 · Data & Measurement on cleaning data prepares meter data, Chapter 5 · Fluids & Thermal on flow measurement provides the water meter balance, and Chapter 19 · Fluids & Thermal on carbon footprints turns energy into emissions.
Exercises and worked solutions
The numbers in the exercises are for practice.
Try to answer first, then open the worked solution.
The main meter of a hospital records 250,000 kWh, and the sum of its submeters is 221,000 kWh. What is the residual in percent, and what is done if the residual in previous months was about 4 percent?
Check my answer
See the worked solution
The residual is kWh, that is, percent, almost three times the usual residual. Intensity and savings figures from the submeters are held back first, and the person responsible for energy and the integrator check for meters with missing data, scale factors, or new loads without meters.
A hospital with a gross floor area of 42,000 m² uses 6,300,000 kWh of electricity a year. What is its electricity IKE, and why can that figure not be compared directly with a hotel’s IKE?
Check my answer
See the worked solution
kWh/m² per year. Hospitals and hotels differ in type of service, operating hours, and ventilation needs, so a meaningful comparison is with similar hospitals or with previous years in the same building.
The mean temperatures on three days are 27.5, 23.0, and 29.5 °C with a base temperature of 24 °C. What is the CDD for those three days?
Check my answer
See the worked solution
The first day is , the second day is 0 because its mean is below the base temperature and CDD is never negative, and the third day is . The total is CDD.
The weekend profile of an office building shows a daytime load of 95 kW and a night load of 120 kW. What deserves investigation?
Check my answer
See the worked solution
The night load is 25 kW higher than the daytime load even though both periods are unoccupied, so there is equipment that runs only at night. The technician finds its hours from the 15-minute profile, matches them with the schedules, BMS trends, and submeters, and reports that equipment so that it can be assessed whether it may be switched off.
A supplier promises savings of 3 percent of the whole building’s electricity and wants to prove it from the PLN meter. What is your response?
Check my answer
See the worked solution
Savings of 3 percent are far below the threshold of about 10 to 20 percent suggested by the FEMP M&V guidelines for option C, so the monthly variation of the building meter can mask them. A more honest proposal is a submeter on the system affected by the measure, with its own baseline before the project starts.
With the model , a month with 140 CDD measures 372,000 kWh. What is the difference from the model, and what does it mean?
Check my answer
See the worked solution
The model predicts kWh, so the measured value is kWh, that is, 2.4 percent below the model. That difference only has meaning if it is larger than the scatter of the model and persists for several months, so it is recorded and monitored, for example with CUSUM as in Chapter 15 · Data & Measurement.
References
External sources
- Metering Best Practices: A Guide to Achieving Utility Resource Efficiency, Release 3.0wbdg.org
- M&V Guidelines: Measurement and Verification for Performance-Based Contracts Version 4.0energy.gov
- Supplement to M&V Guidelines: Measurement and Verification for Performance-Based Contracts Version 4.0energy.gov
- A guide to display energy certificates and advisory reports for public buildingsgov.uk
- Operations & Maintenance Best Practices: A Guide to Achieving Operational Efficiency, Release 3.0energy.gov
Revision history
| Date | Version | Change |
|---|---|---|
| November 13, 2026 | 1 | First published |
How to cite
Kurnia (2026). Building energy dashboards: meters and submeters, energy use intensity, baselines and weather normalization, reading waste from trends, and verifying savings. ARSTEMA Academy. Revised November 13, 2026. arstema.com/en/academy/konsep/dasbor-energi-gedung-meter-dan-submeter-intensitas-konsumsi-energi-garis-dasar-dan-normalisasi-cuaca-membaca-pemborosan-dari-tren-serta-memverifikasi-penghematan
Self-check · 3 questions
Check your understanding before moving on
Write your answer in your own words, then compare it with the worked solution from this chapter’s exercises. The system doesn’t judge open answers as right or wrong; you decide which parts you’ve understood.
The main meter of a hospital records 250,000 kWh, and the sum of its submeters is 221,000 kWh. What is the residual in percent, and what is done if the residual in previous months was about 4 percent?
A hospital with a gross floor area of 42,000 m² uses 6,300,000 kWh of electricity a year. What is its electricity IKE, and why can that figure not be compared directly with a hotel’s IKE?
The mean temperatures on three days are 27.5, 23.0, and 29.5 °C with a base temperature of 24 °C. What is the CDD for those three days?
Nearby chapters
Reading order by shelf: Buildings & Facilities · Chapter 35.
Reading order by shelf: Buildings & Facilities · Chapter 33.