Get your geek on—energy modeling is poised to play a large role in construction decisions, including small-scale residential projects.
Energy-modeling software is more accessible than ever. Good thing—many incentive programs now require scientific validation to “prove” a home’s energy performance before you can cash in. From a builder’s and homeowner’s perspective, and short of actual construction, modeling is the most accurate method of comparing building designs. You get to test-drive your designs and tweak them in response to the results, saving yourself from expensive design disasters and a poorly performing house.
Building simulation computer software can model a home’s year-round energy performance. The goal of a simulation is not only to predict how much energy a building will use, but also to compare alternate designs. These programs also allow you to investigate source-energy consumption and carbon dioxide production. Sophisticated software can even predict the number of hours that the building will be outside of the human comfort zone.
The energy flow (Sankey) diagram illustrates the complexity of energy flows for a typical house. Based upon a building description that is input by the user, the software calculates all the energy flows, including the thermal effects of people, infiltration, lights, and other loads. The flow of energy is further complicated because building materials are not just insulators; they also have a certain heat-retention capacity. Everything from furnace efficiency and fan performance to daylighting controls and heat loss can be simulated on a continual, year-round basis.
With all this data, it’s clear that a simulation can be very helpful with building design decisions. Authorities that regulate energy compliance and agencies that give out rebates want to see this kind of analysis to make sure that buildings qualify.
Quickly and accurately comparing various ways of building, different equipment, and even alternate control strategies, yields a much better chance of optimizing building envelope and systems based upon the actual climate, planned use, and economics.
Besides analyzing the all-important design day, which represents the climatic conditions that yield the highest thermal load (also known as peak load), nonlinear system effects may be revealed—where simple changes yield complex effects. Nonlinear means that many of the house’s energy inputs are not proportional to their loads. For example, if you double the cross-sectional area of all the ductwork, then the (theoretical) fan power decreases to one-quarter of the original. In turn, the fan motor heat is reduced. Even without increasing duct size, if airflow is reduced by half, the software predicts that our fan power should reduce to 1/8 of the original.
Computer simulations examine alternatives such as increasing or decreasing mass, using equipment with better unloading capabilities, and oversizing systems that would otherwise be evaluated only by an educated guess. Before you drive one nail, you can optimize your design, leading to savings in time, labor, materials, equipment, and energy.
To design heating, ventilation, and air-conditioning (HVAC) systems that provide comfort and economy, engineers have always relied on calculations to predict heating and cooling loads. When cheap computing power became available, hand calculations were automated. As computers became more powerful, more sophisticated and accurate algorithms replaced approximation methods.
Commercial simulation software offers a great deal of programming flexibility. With that flexibility comes many choices, and a steep learning curve. Because these tools offer so much potential, there is a lot of activity focused on making the software more accessible by offering graphical interfaces and building model templates.
Simulation software uses the physics of heat transfer combined with environmental conditions to determine internal temperatures and heat flows. This is the calculation “engine,” which can use various load calculation methodologies, with most of the differences usually only of interest to researchers. Although the software is very complex, conceptually there are three main sections:
In this phase, the task is to take all the details of a building and input the parameters into the software. For the building envelope itself, the order of information is:
Then we do the same for interior surfaces, since even with a single-zone model they can affect the building’s thermal performance and capacity.
Because internal loads—such as people, lights, set points, and the schedules that govern them—can also influence the building envelope, they need to be included. To describe these loads, which vary over time, we use the concept of ”schedules.” An example would have inputs for time, day of the week, and holidays so that the software will be accurate any time of the year and determine what the setpoint should be, or how many people are expected in the zone, or any other event you wish to model.
Buildings are divided into thermal zones. A zone is defined as the space controlled by a single temperature controller. Residential systems typically have only one thermal zone, controlled by a single thermostat. This is where the model loses some accuracy, since different rooms in a house can be comparatively cooler or warmer. The justification for this thermal control and modeling inaccuracy is that occupants often move based on comfort. If the bedroom is cold but the sun is shining through the living room windows, just like Fluffy the cat, we can relocate ourselves to that sunny spot on the couch. For modeling purposes, specifying an average temperature is usually good enough.
Once this part of the model is complete, the software can calculate the “loads” by applying the physics of heat transfer, which are incorporated into algorithms for heat conduction, radiation, and convection. The simulation engine churns through millions of such calculations to determine the cooling and heating loads for every hour of the year. Each calculation uses all of the previous inputs, including climate, building envelope, and schedules (occupancy and setpoints) to determine the energy required to maintain the temperature set point—the target value of a controlled parameter. (For example, a room temperature set point would be 70°F for heating and 74°F for cooling.) The weather is simulated by applying a full year of climate data for the home’s location, which can include temperature, humidity, and cloud cover.
Just as your car does not deliver its best fuel efficiency when you’re stomping on the gas, a mechanical system does not deliver the same performance or efficiency under varying conditions. In this phase of the simulation, the software uses the calculated thermal load and the characteristics of the mechanical equipment to determine the equipment’s response (on or off, fan speed, etc.).
The mechanical system’s response to a call for heating or cooling is not always as expected. In a commercial system, a zone call for cooling can set off a chain of events, from increased local air flow to increased speed of a fan, to more cold water from a chiller. To correctly model energy usage, the software has to predict the interactions of all these pieces of equipment as they attempt to do their job. If the equipment has been modeled correctly, we can learn a lot.
Most of the time, equipment operates below its maximum capacity, otherwise known as “part-load” conditions or “partial loading.” Part-load performance can be more important than full-load efficiency in determining annual energy use. An example of how researchers have attempted to document this is the Seasonal Energy Efficiency Ratio (SEER), which gives the efficiency of residential air-conditioning units as a weighted average under different outside air temperatures. But an energy simulation is much more accurate in predicting energy use than such a one-size-fits-all parameter.
The Department of Energy’s free EnergyPlus software is probably the most technically deep software—with a 3,000-page manual, it’s considered the Swiss army knife of simulation software. Plus, it’s free to download.
If you want less of a learning curve, simpler software with fewer building parameters may be a better choice. Many of the packages use a graphical interface, which makes the initial building description phase straightforward. The Home Energy Efficient Design (HEED) software is particularly easy to use, designed for residential applications and incorporates energy-saving strategies into its design alternatives.
Remember that your results will only be as accurate as the data you supply. Inaccurate data in equals inaccurate data out, so make sure you check and double-check your input parameters.
The purpose of the runs outlined in the “Test-Driving EnergyPlus” sidebar was to get an overview of which systems and strategies were worth further investigation. In particular, the hunch was that phase-change material, which is great for load shifting, would work synergistically with night ventilation. Cooling set points were set at 73.4°F. Each case has air-conditioning, however, cases 2 through 5 have the mechanical cooling system disabled from 7:00 p.m. to 12:00 p.m., and depend upon either night ventilation or thermal capacity for cooling at those times.
As expected, the modeling showed that night ventilation via whole-house fans used less energy for each of the selected climates. Increasing the thermal capacity by adding PCMs helped reduce cooling costs. However, the best savings resulted from using PCMs with a large, brushless permanent magnet-motor-equipped fan. Interestingly, using the large whole-house fan (case 4) saved less energy than using a smaller fan. This appears to be due to the fan using nearly as much energy as it saves. A better model using a more realistic control sequence (as a smart homeowner would do, using fans depending on comfort) seems like an interesting area of inquiry.
Each of the runs also provides a “comfort report” showing the number of hours that are outside of the comfort model. There are virtually no uncomfortable hours for the San Jose location; Sacramento cases averaged 120 hours, while the Boston climate reflected up to 180 hours that did not fit within comfort standards. A next step might be to refine the energy models for a particular climate by using more sophisticated controls to optimize comfort and economy by deciding when to use mechanical cooling or night ventilation.
Energy simulations are great tools for building and designing energy-efficient buildings. Creating a basic model can be very quick and effective to address basic building and mechanical system design issues. If time permits, there is even more opportunity to optimize the building envelope, siting, shading, and mechanical systems. Plus, these tools are certainly more powerful and offer much more insight and validation than some “point systems” that are commonly used to determine energy compliance.
Neil Smith (neil at airscapefans.com) is a professional mechanical engineer. Neil’s interests include HVAC and energy efficiency, and he currently runs the AirScape whole-house fan company.