10 Steps To a Useful Energy Model

In today’s world of high-tech buildings, increasing energy prices and ever-more-stringent building codes, designing and maintaining a building is complicated. Many governments (municipal, provincial/state or federal), institutions and companies are realizing the potential benefits of reducing their energy and resource use. From simply saving money, simplifying production lines,  and being environmentally responsible, to marketing and public relations exercises, there are many reasons to want to understand the energy use patterns in your buildings. An energy model is an excellent way to do this and can be applied at any time during the life-cycle of a building. Ideally, modelling would come into play very early in the design phase, but there is still great value to be found later when energy retrofits are being considered.

An energy model is simply a mathematical (software) representation of a physical building – one that already exists or one that is being designed. The model is a tool for simulating what will, could, should or would happen in real life if the exact circumstances of the model happened in reality. This gets complicated very quickly, as obviously it can be very difficult to know what is or will actually happen in real life. Anyone new to building energy modelling might look at a finished model and feel overwhelmed at the shear amount of information that goes into a good model. Like driving a bike or a car, however, once you get the hang of it you won’t feel as intimidated by the details.

I don’t want to claim that there are only 10 things to do in order to create a useful energy model, because sometimes (if not most of the time) there are many more. For someone getting started though, these 10 steps should help clarify the types of tasks, documents, information, tools and other resources that are required to create a useful model of the energy use in a building.

Before I get to the 10 things, what exactly is a useful energy model? Useful is relative so a “useful” model is one that meets the needs of the modeller and fulfils the purpose for which the model was created. This, then, brings us to the first and arguably the most important step:

1. Identifying the Purpose of Your Energy Model:

Before you can take even the first steps, you need to know why you are making an energy model. Are you intending to inform the design process for a new building? Are you modelling an existing building for the purposes of evaluating energy retrofits? Are you trying to judge the relative energy use of a portfolio of buildings? Are you trying to identify inefficiencies in building processes? Are you aiming to achieve LEED certification for your project? There are many uses for energy models and each has specific requirements.

An existing building vs. a new one, for example, requires a very different approach. In each case there will be very different goals with significantly different information available. The decisions about the model’s level of detail, the type of information you gather and the way the model is validated will be different and can required different skills/experience. For projects involving LEED or other building certification programs, the requirements of the program in question will dictate many of the decision made during modelling.

2. Information Gathering and Review:

This stage can be relatively straight forward if the project is under design (get the most recent set of drawings) or very time consuming for a large existing building with no documentation (a week-long site audit). Generally, though, the same basic information will be required. This includes, but is not limited to:

  • The building’s shape, size and other geometry;
  • The building’s construction materials, insulation, windows, etc.;
  • The use and schedule of the building and its systems;
  • The control sequences, methods and intentions;
  • The location, climate and site orientation, and;
  • Lighting, equipment and process information.

This documentation can come in the form of drawings, Operation & Maintenance manuals, interviews with site personnel, equipment cut-sheets, climate databases, and any other documents typically associated with buildings and their operation and/or construction. For an existing building, much of this information will be gathered on site during an audit. However, some of it will inevitably conflict with what is observed during the visit. This requires detective work and personnel interviews to determine what is the current state of the building.

The review of this information is also critical. It is important, for example, to have a good understanding of the design methodology of the HVAC systems in the building before attempting to model them. Are the systems air-based or hydronic? Are they CAV or VAV boxes in the ductwork? Is that rusted-out unit on the roof actually in operation? Will the intended design include an air curtain? These and many other questions, applicable only to the project at hand, need to be understood. This process takes experience to get right and even those who have modelled many buildings will find themselves part-way through a model when they come across information that didn’t gather at the beginning or that they didn’t realize was needed.

3. Base Year Selection and Analysis:

The topic of base year selection can’t be summarized easily and deserves its own post, but the basic idea, though, is to select a representative year of weather data to reliably validate your energy model. This is typically done is by comparing the predicted energy use of the model to the base year and correcting the model until a suitable match is achieved. A base year will generally apply only to existing building modelling.

A year of weather data is representative if it can be considered to be “typical” and if it includes all of the seasonal weather variations, scheduling changes and other occupancy and operational changes that are unique to a building. If a major renovation or occupancy change occurred, it would be best to choose a different base year if possible, or at least to account for the difference during modelling.

This step does not necessarily need to be completed at the beginning of a modelling project, but is a logical task to have done in parallel with the rest modelling process, perhaps by another staff member. The information acquired in this step is needed only at the end during model validation and verification, but it can also bring out important questions about the building’s behaviour and potentially help focus the ongoing review of the building’s systems and how they are performing.

4. Preliminary Brainstorming and Other Planning:

By this point in the process, the modeller should have a good idea of their goals and how they want to achieve them. For example, he or she may know that they will be focusing on a specific wing of a large building due to its abnormally high energy use or that the client for the newly designed building wants to emphasise the green roof feature in their project. This type of planning will inform the rest of the model and some decisions made later.

For a project involving an existing building, this is the point where a preliminary brainstorm of energy conservation measures would be created. It is important to have at least a preliminary list at this point in order to save time on some areas of the model and to focus more on others. These early measures will usually evolve as the project progresses, but it is useful to get started early.

Another task that should be completed by this time is the selection of a modelling tool. In some cases, a modeller may have a “usual” tool they use for most projects, but no one tools does everything so it is useful to have a “toolbelt” of software at your disposal. Is this a low-budget, quick-and-dirty estimate? Is this a multi-million dollar LEED sky-scraper? The tools selected in these cases will most likely not be the same. The topic of which tools to use when also deserves a post of its own at some point in the future.

5. Preliminary Zoning of Building for Take-offs:

Before the building’s information can be entered into the modelling software, the building should be “zoned” or broken down into distinct areas. If a large, multi-floored building is modelled as a single zone (the equivalent of assuming that it’s a single large enclosed space), the accuracy and usefulness of the resulting model will be low. On the other hand, modelling each and every room in a large building is time prohibitive, not to mention the time it would take the modelling tool to simulate every room. So there is a balance that needs to be struck in selecting the appropriate level of modelling detail – this takes experience to get right and is as much an art as a science!

As a rule of thumb a space should be “zoned” separately if:

  • it is served by a distinct HVAC system;
  • it has significantly different occupancy or temperature setpoints;
  • it has significantly different building envelope or insulation values;
  • it will be the target of a specific, known energy conservation measure;
  • it contains significantly different lighting or equipment loads, or
  • the area was or will be altered significantly in the near future.

6. Geometry and Physical Characteristics (Take-offs):

This stage involves acquiring and then entering the geometry and building construction data into the chosen modelling tool. This is an ideal step to have performed by someone other than the primary modeller in parallel with other modelling tasks. The exact details of what would be involved differ greatly from project to project and between different modelling tools, but generally involves reading information off drawings or other documents and then entering wall, window and door dimensions into the chosen tool. Also entered at this point would be the wall, window and door constructions and the properties of any glazings if available.

With the rise of BIM systems, this process will become increasingly streamlined, at least for new buildings, but will remain time consuming for the foreseeable future when it comes to existing buildings, especially those with few electronic drawings.

7. Primary Modelling:

Here we are at Step 7 already and only now mentioning actual modelling! This is worth noting because it shows that the preparation and foundational stages of the modelling process really do make the rest of the process easier. In this stage the actual modelling (in the expected sense or using a modelling tool) takes place. Heating, ventilation, air-conditioning and other building systems are entered into the modelling software along with lighting systems, domestic hot water systems, and any plug- and process-loads.

An important part of this stage, which could be its own separate step, is scheduling. Every fan, pump and light in the building operates according to some schedule, and this schedule has significant impact on the associated energy use. There is also a great deal of scheduling involved in any control systems present. Every valve, damper, actuator, temperature setpoint and more can be scheduled, so this needs to be accounted for in the model.

8. Energy Model Calibration/Validation and Review:

Once the model is “done” (a model often keeps evolving so it’s hard to say when it’s actually done) it will need to be calibrated/validated. In the context of a model for an existing building, this means to demonstrate that the model correctly predicts what actually happens in real life. To do this, the modelled energy use is compared to the actual energy use for the base year (chosen earlier). A monthly comparison is desirable, but at least the annual total should match relatively well.  There are some standards which try to define “well”, but the degree to which the modelled and actual data should match is debatable and depends, again, on the purposes of the model.

If differences are found between the modelled and actual data, the model will need to be corrected (assuming the actual data is take to be valid, which is not always the case). If the difference appears in the winter only, for example, there may be an error involving the heating system or some season schedule. This process involves a lot of detective work and requires significant knowledge of the model, the tools being used, the building in question as well as building systems in general.

If the model is being used for a new design, this step may not occur until after the building is constructed (when there is actual data to use). In either case, though, a review of the model for quality control is also appropriate at this stage. This review could involve another modeller who would review key variables or a check-list used by the main modeller him or herself. In larger companies, there is a standardized method to model specific types of systems and certain quality control processes in place to ensure the standards are followed. This type of quality control system may not be applicable to a small company or project, but it is always useful to do some form of “sanity check” on the inputs and outputs of a model.

9. Use the Energy Model for its Purpose:

The model can now used for its purpose, whether that be savings calculations for an energy conservation measure, or making an estimate of the impact of the latest design alterations. It’s important to note that no model (or modeller) is perfect and all outputs from any model should be fully understood before being handed over to a client or used as inputs to another calculation. Ultimately, it is the engineer, technician or other professional who is responsible for the results of a model – not the modelling software! This is where experience comes in. An experienced modeller needs to be able to see, for example, that the distribution of energy end-uses is not right for a given building, or that the energy density for a specific area is too low, etc. Some projects are underway help this by creating benchmarking tools of typical buildings, but the final decision will always rest with the professional.

10. Iterate as Required:

Throughout Steps 7, 8 and 9, errors, omissions or inconsistencies may be found in the model. Also, the particular way in which certain aspects of the building were modelled may be found to be inaccurate. If any of these occur, changes will be made to the model and outputs may need to be recalculated. The modelling process is very much an iterative one, gradually improving the model, adding detail, if required, until the final results are achieved.

Iterations may also be required as the design process evolves. In early stages a simple model may be enough, but later a more detailed model, possibly a new model made using a new tool could be necessary.

Final Comments and What’s Next

It should be noted that many of these steps can be done in parallel, while some must be done in order. It’s perfectly feasible to have someone working on the base year while a second person is doing architectural take-offs while the primary modeller reviews drawings to being understanding the building. However, the primary modelling really can’t take place until after the geometry and building construction is entered into the tool.

These steps are also greatly dependant on the purpose for the model and the tools being used. Some tools, for example allow importing CAD drawings, side-stepping a significant portion of the geometry entry. The amount of time and effort spent on each step will also vary according the model’s purpose. If the energy consumed by an ventilation fan on an interior server room is the only point of interest, it may not be necessary to model any significant part of the building envelope at all.

Other energy modelling resources include:

  • OneBuilding.org: This site includes various useful discussion lists regarding energy modelling.
  • DOE Software Tools List: This database of energy-related modelling tools is maintained by the US Department of Energy.

It’s obvious that energy modelling is a huge topic with lots to understand, and that it’s as much an art as a science, so like many things it takes practice! Each of the steps discussed here could easily be expanded on into entire books of their own, so I intent to touch on many of them in greater detail in the future. Please stay tuned and if you are particularly interested in any step in particular, leave a comment below to let me know!

This entry was posted in Energy Modelling on by .

About Matt

Matt is a mechanical/building engineer who specializes in whole-building energy modelling, energy efficiency and solar buildings. He's worked for about 9 years analyzing the energy use patterns of buildings. He studied the energy performance of a low-energy, solar house for his Masters thesis. Matt has experience with EnergyPlus and eQuest.

Leave a Reply