Skip to Content
U.S. Department of Energy Energy Efficiency and Renewable Energy

CREEA Whole Buildings Baseline Webinar (text version)

Below is the text version of the webinar titled "CREEA Whole Buildings Baseline," originally presented on May 20, 2011. In addition to this text version of the audio, you can view the presentation slides and a recording of the Webinar (WMV 57 MB).

Announcements
Presenter: Doug Black, Lawrence Berkeley National Laboratory (LBNL)

Black announced that he would mute all during the webinar and would then open for questions at end of the webinar, which is scheduled to run 30 to 40 minutes. Black also announced that this webinar was being recorded.

Black discussed the focus of this webinar: What is an energy baseline? When do you use one? Why is it important? The inputs, data, to create a baseline, and where to find resources and tools that can help.

What is an energy baseline?
With a goal to reduce energy consumption and carbon footprint, how do you decide from the many energy-efficiency measures (EEMs) which ones should you implement? How do you financially and operationally justify selection? After you have selected and implemented measures, how do you know that it works, or how much have you saved to quantify return on investment? The key to answering these questions is an energy baseline. This is where you start; this is a representation of the energy consumption of a building in the form of a calibrated model or a set of energy end-use measurements. A baseline represents building energy use in its current state with either a calibrated energy simulation model or with measured energy data.

Where does an energy baseline fit?
Benchmarking (e.g., ENERGY STAR® Portfolio Manager) along with baseline modeling are ways to identify savings. Distinctions between benchmarking and baselines lie in that benchmarking is coarser, relies on comparisons to peer groups where there may be variations in those groups, provides broader guidance on approaches, can be less complex, and is typically applied to cases with less system interdependence. A baseline energy model is specific to the building and can provide detailed guidance in energy-savings measures, especially with capturing effects of interactions of measures in different systems. Of course, with greater capability comes greater cost, compared to benchmarking.

Why create energy baseline?
Without it, you cannot quantitatively evaluate an array of measures, especially evaluating the interaction of multiple EEMs. You cannot measure performance of EEMs. Also, with greenhouse management reporting being increasingly required, it is critical to have a baseline as well as for voluntary commitments or programs like ENERGY STAR and the Better Buildings challenge that is ramping up, and also potentially with mandatory reporting as part of sale of commercial buildings. You may want to create an energy baseline as part of your management plan.

Examples of EEMs Suitable for Benchmarking Analysis
To stress the distinction between benchmarking and a baseline, benchmarking analysis is better suited to evaluate energy-efficiency measures that are well defined and better understood in industry in terms of impact (outside air economizers, etc.).

Examples of EEMs Suitable for Energy Baseline Models
Baseline energy models are better suited to evaluate measures that are more complex where you need to capture effects on multiple systems (large building tenant improvement, etc.).

Example of an Energy Baseline
Black showed a very simple baseline output. The sample shows monthly electricity use for a reference ENERGY PLUS model for a large commercial building in Chicago. Blue bars show baseline electric energy use predicted for each month. The other bars show model predictions when two different EEMs, EEM1 and EEM2, are implemented to illustrate how a baseline model can be used to select between two different approaches. Another benefit of a baseline model is shown in the pie chart, which shows a detailed breakdown of energy use.

Steps to Creating an Energy Baseline
The first step is to set goals or objectives, prioritize many potential EEMs, etc. The first objective requires a detailed energy simulation model; other objectives can be accomplished with a model and also with energy-use data and relatively simple analyses. The key is the collection of building system and component details and data. This is the focus of this webinar: what data you need to collect to create and calibrate a baseline energy model. Black summarized what needs to be collected and said that details will be provided later in the webinar.

Building Energy Modeling
These are sophisticated programs (ENERGY PLUS, eQUEST, DOE-2) to predict performance of single and especially multiple energy measures, and identifying the best "bang for the buck." You can use these models to identify energy savings with changes in sequences or set points and staging. To increase confidence, models need to be calibrated with actual energy data—the more the better. Black cautioned: A calibrated model can predict well for one set of conditions, but not for others, e.g., for a set of conditions for one period, but not, for example, extreme weather. Models are only as good as the input and expertise of the modelers. This can be relatively expensive for large detailed projects.

Energy Model Input—eQUEST Construction Materials Inputs
Black gave details of inputs that go into these models and showed an example eQUEST input interface. The text portion in the upper table shows how different construction layers are defined—walls, floors, roofs are each defined as layers—and for each of those materials the model needs specific thermal property data. You only need to provide wall construction materials from construction drawings. The modeler will use this information to define thermal parameters and input those into the model.

Energy Model Input—ENERGY PLUS HVAC Fan Inputs
Black showed an ENERGY PLUS screen shot showing variable volume fans, with detailed parameters to define coefficients. You must provide the make and model of the fan to determine the specific coefficients for the model.

Black then discussed specific inputs to provide to the modeler:

Building Properties
From construction drawings: dimensions, occupancy, functions, wall construction, surfaces, materials, window sizes, window-to-wall ratio, specifics about the window type, make and model to determine light transmission, window covering and shades, insulation of exterior surfaces, insulation U-values, installed lighting and lighting control, occupancy controls, and plug loads are usually estimated, but significant ones should be recorded. Sources for information can be from as-built construction drawings.

HVAC System Properties
The key things are the components and make and model of chillers, boilers, pumps, etc. You must compile the control sequences for all HVAC components. Sources: as-built controls drawings, sequences, and submittals.

Operation Schedules and Set Points:
Set points and schedules are also key. They can come from the building management system (BMS): zone set points, economizer set points, lighting schedules, general occupancy schedules, etc.

Weather Data
Weather has a big impact on energy performance and on model predictions of energy performance. To have actual historical site weather data is really valuable, especially for weather-dependent energy-savings measures. Turn on trends of any weather measurements made as part of a BMS such as dry-bulb temperature, wet-bulb temperature, RH, etc. Another source can be local weather data, e.g., www.wunderground.com. If you're considering natural ventilation as a potential retrofit, install a good anemometer to store wind speed as far in advance as possible; it is a good worthwhile investment. It is good to have site-specific data for this type of analysis.

Whole-Building Utility Meter Data
Data is needed to calibrate an energy model or to use in a baseline analysis. Whole-building energy data is usually available from utility companies; ENERGY STAR Portfolio user utility data can be used. Time resolution may be hourly or at 15-minute intervals. If you have multiple meters, you will want to specify which system or spaces are connected to which meter. If you have district chilled water, heating water, or steam, it is good to have meters on these.

Submeter Energy Data
In addition to the whole-building level, it is good to have submeter data, especially for major system retrofit projects that involve multiple systems, such as those involving HVAC or the building envelope. Sometimes this is not difficult to get with the increasing use of variable frequency drives. Often, the power data is available at controls but may not be tied to the management system. In these cases, you need a simple data logger to record output at drive. Data loggers are fairly inexpensive, do not require pulling wire, and are easy to operate, but they require labor to download, maybe once a week or every few weeks. If the power data is not available from the controller, you may want to put a power meter on the device. Also, measuring the energy of chilled and hot water systems requires the use of Btu meters or separate measures of supply and return temperatures and loop flow rate to calculate energy loads. The caution here is that these meters can be somewhat expensive. If submetering can become part of a continued monitoring system, it would be beneficial to operations and performance maintenance. If permanent installation is not possible, then the duration of submetering depends on the system that you are monitoring; operating conditions, for example, with a lighting retrofit, might need as little as two weeks to sample, but chiller-related retrofits might need six months or more.

Building Management System Data
There are other ways to get energy data from building management systems. Constant-speed motors will report status. Turn on trends from constant- and variable-speed motor controllers to trend on/off status for constant motors and speed for variable motors. These points can be converted to energy use with other information, such as motor-rated power and performance curves. There is some error involved in this, but it is a good proxy. A better proxy is to make spot measurement of power consumption to create a calibration curve for converting the proxy variable to a power value. It only takes a one-time measurement period. This can be useful. Other methods are flow rates of chilled and hot water, supply return temperature, and air temperature. Turn on trends in building energy management systems to store measured data in the control system.

Data Management

Pulling information together can be challenging. It can be kept in binders or file drawers, as data generated can be a significant amount. Databases with tabs can be used. Data from different sources with different time periods, meters, systems, and trends can be challenging to analyze. Missing and erroneous data can be averaged, but sections may need to be taken out of analysis. A good tool to deal with varying time intervals, time step formats, and missing data is the Universal Translator (utonline.org).

The Value of Data
Computer memory is inexpensive. Compiling information and setting up systems to trend can have costs involved, but this is inexpensive compared to value in historical data. The power of data to be used to save energy and to prioritize and verify savings and maintain is very valuable.

Example of Regression Analysis
Some simple tools to use can be very valuable. This slide shows simple regression that refers to putting an average line; the example shows whole-building daily electric use, with plotting showing before and after energy-efficiency measures were put in place. The sample comes from a real-world case. The scattering data with the line drawn shows less energy use. This is pretty simple to do in Excel with the trend line feature.

Normalizing Energy Consumption
Another way to normalize data is to take into account cooling degree days and heating degree days. To compare different time periods, you need to take into account the weather differences. CDDs and HDDs are available from degreedays.net A simple example was shown.

Summary
Black summarized his points: An energy baseline characterizes the current state of a building's energy consumption. A powerful tool for creating a baseline is the calibrated energy model, which can be completed with measured data and simple analysis tools. The key to the creation of an energy baseline is to get various building data and energy end-use data. Combined with a good energy modeler, the data can identify and help prioritize energy measures.

Final Thoughts
Getting energy data can be a big exercise in compiling and organizing large amounts of data. Energy modeling is a powerful tool, particularly to quantify multiple energy-efficiency measures. The effort involved can be great and is most valuable for large retrofit projects.

Building and HVAC Model Inputs
Tables show building information needed; the information in parentheses are the make and model. These are critical inputs to modeling.

Resources
These are more for energy modelers, showing what goes into making a good energy model. Accreditations of modelers was shown. Other resources and tools were presented.

Black opened the webinar to questions from the attendees.

Question: Do you have any guidance on modeling how to find and retain someone with appropriate expertise, what guidance might be out there to determine which modeling tool might be appropriate for a particular situation?
Answer: Efforts are being made to create accreditation for modelers, but that does not yet exist. In the national labs' role with the Commercial Building Energy Alliances (CBEAs), we can't recommend vendors. Try word-of-mouth and ask around or work with people that you know are good from related projects. The level of modeling depends on how much system interconnectedness is involved; for less complex measures, you want to use simpler analysis tools like regression. Does Cindy Regnier have a response? I can't offer a whole lot more, but finding energy modelers who have demonstrated expertise shows that they are serious about energy modeling. Tools have strengths and weaknesses; it is really kind of in the weeds, unfortunately. Some tools are suited for quick and easy, while others are more sophisticated. The question is complex, unfortunately.

Question: What is a reasonable expectation for a calibrated model predicting building performance?
Answer: With experience, getting within 10% is really good. This should be the target. Even better for 5% effect. Models that capture a lot of specific energy-use data capture weather variations monthly can do it at 10%. Cindy Regnier added that it also greatly depends on the accuracy of measuring devices; accuracy costs more. It also comes down to what kind of granularity is important for measures being looked at. You can do better than 5%, but this comes with cost. It depends upon if it is a question of decreasing loads to attempt removing a chiller, you need greater modeling accuracy. A lot of this information also relates to verification of implemented measures. You need a modeler on board. A good measurement and verification contractor can also help with advice on what are good value propositions for selecting measuring devices.

Question: Have you ever attempted to measure synergistic effects of multiple measures? Answer: Individual measures have certain savings, but when you combine several together, sometimes they don't add up. Regnier said this is known as a parametric analysis. A modeler will identify measures. One way to find a sweet spot of energy-savings benefit is to bucket into suites of energy-efficiency measures and perform incremental runs. It can be a lengthy process; someone with experience may help define the buckets that are of interest to the project. The combined impact depends on measures. Professional judgment is needed.

Question: For large facility retrofits, can you ballpark cost?
Answer: Regnier said it depends. One really good tactic to reduce costs is to look for the ability to replicate in building. Hopefully, you can condense it down to a typical floor configuration. Also, you want to pay attention to usage patterns. Pick something representative; you will have outliers, so try to condense the effort to a degree that makes sense. The cost depends on the size and complexity of the building.

Question: What is the cost for a pretty straightforward, owner-occupied building?
Answer: Regnier said it could be in the tens of thousands of dollars. A robust and large building can be breaking the $100,000 barrier, or even more. The listener found the answer helpful. Regnier said that some utilities offer incentives. Check for such a program in your area.

Regnier announced a submetering webinar on June 16 and a retrocommissioning webinar on July 14. These relate to the topics discussed in this webinar.

Black thanked the participants and ended the webinar.