UPDATED February 1, 2012
In 2004, researchers at the National Renewable Energy Laboratory (NREL) developed BEopt, a software program that finds the least-cost solution to designing a zero-energy house. Now that the software developers — a team that includes Craig Christensen and Scott Horowitz — have spent seven years improving the program, it has finally been released to the public. The development of BEopt was funded by the U.S. Department of Energy.
BEopt can be downloaded for free from a new BEopt website maintained by NREL. In addition to its original functions for designers of new homes, BEopt now includes new functions that prioritize energy-retrofit work in existing homes.
The BEopt website includes links to “help” files and training videos to get new users oriented. Although the software still has a few flaws and limitations, patient designers will find BEopt to be a useful and intriguing software program.
How good does a home’s thermal envelope need to be?
BEopt performs at least two functions: it is an energy modeling program that calculates the expected annual energy use for any house design, and it is an optimization program that identifies the least expensive way to build the envelope of a net-zero energy house. Users of BEopt can access weather files for 1,000 locations in the U.S.
The program includes a database of user-modifiable construction cost data. This feature allows BEopt to determine (for example) whether a house with 12-inch-thick double-stud walls and double-glazed windows will cost more or less to build than a house with 2-inch-thick foam sheathing, 2×4 walls, and triple-glazed windows. It will also determine which of these options will perform best.
BEopt shows designers of zero-energy homes how to improve envelope specifications, and lets them know the right point to quit making envelope improvements. After all, including energy-efficiency measures in a new home reduces the total cost to the homeowner for utilities and mortgage payments compared to a code-minimum house — up to a point. As more and more improvements are included in the design specifications, energy costs get lower and lower, until a point is reached that is optimal from a cost perspective. Beyond this optimal point, the designer may be able to achieve further energy savings, but the monthly costs to the homeowner for utilities and the mortgage begin to rise.
It’s a useful program, even if you aren’t building a zero-energy house
Once envelope improvements start costing more than a photovoltaic (PV) array, they become a poor investment. Most homeowners don’t want to pay $5,000 for extra insulation if the investment yields lower annual savings than a $5,000 PV system. By identifying the point at which envelope improvements stop making sense, BEopt helps designers arrive at the least-cost envelope for a zero-energy house.
Even if you aren’t interested in including a PV system, BEopt can still be useful. After all, it always makes sense to design a house with the lowest combination of utility bills and mortgage payments — even if it isn’t a zero-energy house.
If you assume that energy prices are likely to increase steeply in the future, you can input your assumptions about energy price inflation into BEopt. It’s possible to enter different rates of inflation for four different types of fuel: electricity, natural gas, propane, and fuel oil. Designers who assume energy prices will increase steeply will end up designing homes with more insulation than designers who assume that energy prices will stay steady.
The software engine is DOE-2
A paper authored by Ren Anderson, Craig Christensen, and Scott Horowitz describes how BEopt works: “The sequential search approach used by the analysis method involves searching all categories (wall type, ceiling type, window glass type, HVAC type, etc.) for the most cost-effective combination at each sequential point along the path to zero net energy. Starting with the base-case building, simulations are performed to evaluate all available options for improvement (one at a time) in the building envelope and equipment. Based on the results, the most cost-effective combination is selected as an optimal point on the path and put into a new building description.”
BEopt uses existing energy software programs to perform its energy-modeling calculations. Most BEopt users configure the program to use DOE-2 software for energy modeling and TRNSYS software for calculating the output of a PV array or solar thermal system. (As an alternative, BEopt users can configure BEopt to run EnergyPlus energy-modeling software instead of DOE-2 and TRNSYS.)
BEopt’s energy modeling is based on detailed hour-by-hour simulations. The software accounts for interactions between different building components. For example, if the specified solar heat-gain coefficient (SHGC) of a home’s windows is lowered and the SEER rating of its air conditioner is raised, the cooling energy savings resulting from these two changes is not additive. The software is sophisticated enough to consider interactions between all of the specified measures and to accurately report the effects of these interactions.
Users need to edit the default cost data
BEopt’s energy modeling software includes data on typical occupant behavior, including assumptions about daily hot water usage and the amount of electricity used for plug loads and appliances. These assumptions are based on a protocol developed by Building America program researchers.
Before running BEopt, it’s essential to look over the default numbers in the BEopt “library” of construction cost data. The default numbers won’t work everywhere, since construction costs vary widely across the U.S. Moreover, some costs have changed significantly since the BEopt program was written.
One unit cost is seriously outdated: BEopt’s default cost for an installed photovoltaic system is $7.50 per watt. These days, $4.50 per watt is closer to the mark in much of the country. The price of PV matters, since it’s harder to justify an investment in “heroic” levels of insulation as PV prices drop. Fortunately, it’s fairly simple for BEopt users to edit the library as needed to reflect local construction costs.
Default wall assembly and heating equipment choices are limited
Most users will also need to edit the list of available wall assemblies and heating equipment options. For example, BEopt’s space-heating options do not include furnaces or boilers that burn firewood or wood pellets.
Another limitation: BEopt’s available wall assembly options don’t include any wall with more than 2 inches of exterior rigid foam. If you want to model a building with 4 inches of exterior rigid foam, you’ll have to enter a new option. To do that, right-click on one of the options in the relevant category. This will bring you to a new screen, the “Options Editor,” where you can describe your custom wall assembly and assign an R-value to the new assembly.
Three input screens
BEopt has three input screens — “geometry,” “options,” and “site” — and one output screen.
On the geometry screen, a user enters information on the shape of the house and its orientation. It’s also possible to indicate the location of neighboring houses that can affect shading.
On the options screen, the user inputs the wall specifications (include wall insulation details), the ceiling insulation type and thickness, the glazing areas, the glazing specifications, the HVAC system specifications, the appliance specifications, and several other similar inputs.
On the site screen, the user enters the geographical location of the building, the mortgage period, the mortgage interest rate, local fuel costs, the assumed inflation rate for fuel costs, and the analysis period. While the default analysis period is 30 years, a much longer analysis period is possible and fruitful. For example, many Passivhaus designers note that the typical PV array won’t last as long as wall insulation. That fact can be reflected in a BEopt analysis.
Every measure under consideration has a presumed lifetime; if you choose an analysis period of 100 years, and a PV system lifetime of 25 years, BEopt will assume that four PV systems will need to be purchased during the analysis period. If you choose to assume that wall insulation will last 100 years, then BEopt will consider the longevity of the insulation in its economic analysis.
The output screen
The output screen includes three boxes. One box includes a line graph. The line graph shows the percentage of energy savings on the X axis and monthly homeowner costs (utility bills plus the portion of mortgage payments associated with energy-efficiency upgrades) on the Y axis.
If you want to run BEopt in “optimization” mode, you must first instruct BEopt to consider a range of specifications. BEopt then runs several iterations — in many cases, many dozens of iterations — with each iteration based on a different combination of specifications. Each iteration is displayed on the graph as a dot. The lowest-cost dot represents the “sweet spot” — the maximum energy savings achievable before the cost curve starts heading up. At that point, if a designer wants to continue lowering the building’s annual energy costs, PV modules can be added to the roof.
The next box includes two bar graphs, with the Y axis showing energy use in MBtu per year. The first bar graph represents the energy use of the home as designed by the software user, while the second bar graph shows the energy use of a home with optimized specifications.
The third box shows the specifications of the envelope measures of the iteration highlighted by the user.
Craig Christensen, the BEopt team manager, notes that a wise designer will look at a variety of solutions proposed by BEopt. “The program allows the designer to look at several design alternatives that are equivalent in cost and performance,” he says. “It allows you to choose what you like best. We don’t just say, ‘This is THE design.’ The program presents a set of design alternatives.”
The program has a few limitations
BEopt still includes a few quirks:
- All walls in any modeled house must have the same characteristics. For example, BEopt won’t allow a user to model a house with CMU walls on the first floor and wood-framed walls on the second floor.
- BEopt always places windows at the same height; users can’t lower a window’s sill height or raise a window’s head height. Although BEopt lets users input a building’s footprint using the mouse, the program won’t let users to adjust a window’s size or location with a mouse. The only way to adjust window area and orientation is by entering the number of square feet or percentage of total window area on each orientation. In one of the houses I was modeling, BEopt insisted on putting all of the gable-end windows on the first floor and no windows on the second floor. I discovered no way to override this quirk.
- Some house shapes are impossible to model in BEopt. For some reason, BEopt won’t allow users to draw a Cape Cod house with 4-foot kneewalls located above the first-floor exterior walls. Nor does it allow users to specify an 8-foot-high ceiling on the second floor of a Cape if the area of the ceiling is less than the area of the first floor. Even Craig Christensen from NREL couldn’t discover a way to model a Cape Cod house.
- It’s a little tricky to figure out how to run the program in optimization mode. Here’s the key: after creating the house you want to model, right-click on the tab for the “case” and choose “new case.” A new tab will appear with your new case. In the “Analysis” box at the top of the screen, make sure you are in “Optimization” mode, not “Design” mode. On the Options screen of the new case, highlight a range of specifications rather than a single specification for building components like walls, ceilings, and windows. You can do this by clicking one wall option, then holding down the “Shift” key, and then clicking another wall option farther down the list. That will highlight a range of options. BEopt will consider each of the highlighted options as worthy of consideration. Options that aren’t highlighted won’t be considered. Once you’ve selected these ranges for all relevant categories of building envelope specifications, click “Run.” Beware: since it will take between one and four hours for your computer to complete a single optimization run, you should probably plan to run BEopt overnight.
When users download the free BEopt software, the necessary energy-modeling software is missing. In order to successfully run BEopt, users have to also download DOE-2 onto the same computer. DOE-2 is available as a free download, but the downloading process is cumbersome.
There are three steps required to obtain the DOE-2 software:
- First, download the DOE-2 Licence Agreement Form, print it, sign it, scan it, and send it by e-mail to Jeff Hirsch in Camarillo, California.
- Wait for Jeff to send you an e-mail with the secret password. (This wait can range from just a few minutes to about a week.)
- Use the secret password to download DOE-2. Note: the Web page that Jeff Hirsch’s e-mail sends you to is extremely confusing. It is a list of several dozen files with titles that appear to be gibberish. What you are looking for is the file that includes the code “47g.”
To download BEopt, visit the BEopt website.
A new version is released
On January 31, 2012, NREL announced the release of a new, updated version of BEopt (version 1.2). The latest version includes several new features, including the ability to model:
- Heat-pump water heaters (BEoptE+ only)
- Ground-source heat pumps
- Dehumidifiers (BEoptE+ only)
- HVAC equipment sizing consistent with ACCA Manuals J and S
- Insulation installation derating based on RESNET
- Degradation factors for existing installed equipment
- Simplified inputs for air conditioners and air-source heat pumps
- Attached walls for simple geometries
- Outdoor air temperature reset controls for boilers
Last week’s blog: “Scary Stories for Halloween.”