A computerized energy model is an essential element of building design for high-performance homes, because it allows the designer to predict the energy performance of a building based on specific site characteristics, structural assemblies, mechanical efficiencies, and renewable technologies. By adjusting these elements of building performance, the designer can choose the most cost-effective combination of features. While models are just predictions — and therefore will never be completely accurate — skillful use of a modeling tool should lead to more affordable, energy efficient buildings.

That leaves two questions. Just how important is energy modeling for cost effectively designing zero energy homes? And how accurate is energy modeling?


Now More Than Ever

For several decades, the mantra of high-performance building design was to slash the amount of energy used for the two largest building loads: space heating and water heating. As a result, less energy needs to be purchased or generated on-site. While load reduction is still an essential design goal, the dramatic decline in the price of rooftop solar electric systems creates a new design question: What is the optimal balance among shell measures (insulating and sealing walls, windows, ceilings, etc.), high-efficiency mechanical equipment, and on-site renewable systems? However, there are four factors that come into play in different projects and regions of the country that make answering the question more complicated:

  • A wide variety of interwoven energy measures can be taken to reach net zero
  • The cost of renewable, especially solar, energy systems is dropping rapidly
  • Government and utility financial incentives vary in different parts of the country
  • Local equipment and construction costs change rapidly with local market conditions

Instead of making a judgement call based on intuition, energy models offer designers quantitative, project specific feedback and guidance.

The majority of residential buildings fail to take advantage of this tool and rely, instead, on building codes as the sole benchmark of energy performance. Even most third-party certification programs, such as ENERGY STAR, Zero Energy Ready Homes, Living Building Challenge, and Earth Advantage, use energy modeling only for program compliance rather than as an iterative design tool to find the most cost effective set of energy saving and renewable energy measures for a given project. Designers and energy consultants can unlock the value of energy modeling as a design tool by modeling different combinations of energy-related features to see which combinations yield the best performance for the least cost. If your goal is to build an affordable zero energy home, then an energy model is an essential planning tool for making the smart decisions needed to design a  a cost-effective building.


Modeling Programs

There are a variety energy modeling programs to choose from. Since computerized predictions utilizing differing assumptions can never be precisely right, I decided to use two different modeling programs and compare their results with the actual energy usage of my own home. I selected REM/Rate, software commonly used in ENERGY STAR and many utility programs, and a free online calculator created by Build Equinox called Zero. Making this comparison gave me an opportunity to explore options for designing zero energy homes more cost effectively based on the information from these two models.

My zero energy home has 939 square feet of living space and was built in 2015. It’s super insulated, very airtight, uses a mini-split heat pump for heating and cooling. Hot water is supplied by a heat pump water heater. There is a 4.3 KW photovoltaic system on the roof. My home is in Climate Zone 5 with about 7,000 heating degree days. In order to get the most consistent results from the two programs, I used the same inputs for both models.


Modeling Results

The chart below shows results from each energy model compared to the actual energy use of my home.


A Comparison of an Average of Two Years of Actuals with Each Model

Actual Energy Use Model Compared to Actual
2-year Avg
from Actual
Build Equinox Zero
from Actual
Energy Used 3190 4266 3,728 6,232 67% 5,121 37%
Energy Generated (PV) 5460 5356 5,408 5,704 5% 5,991 11%
Net 2270 1090 1,680 -528 -131% 870 -48%

How does modeling compare to reality?

Neither program closely predicted the performance of the home over these two years. Build Equinox Zero predicted energy use that was 37% above the actual, while REM/Rate’s number was even higher. That’s not surprising. Each year presented a widely different weather pattern. For example, 2016 was relatively mild, while the early months of 2017 were marked by sub-zero temperatures and about 30 inches of snowfall that completely covered the solar panels for several weeks. Computer models work with long term weather averages, and so can’t be expected to match any two-year snapshot. To complicate matters, the climate may be changing faster than predicted.

Furthermore, neither computer model can predict residents’ actual energy-use patterns. In the case of my family, we are probably more frugal in our use of energy that the “average” family assumed in these models, which could in part account for both models over estimating our actual energy use. As in the models, a designer or consultant would certainly want to take a conservative approach. Better-than-expected performance would mean that extra energy could help power an electric vehicle or electric yard equipment.

Of the two programs, Build Equinox Zero came closer to predicting actual overall performance in our case. It’s estimate fell into the positive energy category, matching the reality of our home energy use, while REM/Rate was more conservative and predicted that the home would not reach a zero energy level of performance.

Why was neither program very close to reality? To answer that question, it helps to dig into the more detailed reports.

Build Equinox Zero
Heating 662 634 4%
Cooling 159 16 894%
Water Heating 979 967 1%
Plug Loads, Lights & Appliances 4432 3504 26%
Total Energy Use Projected 6232 5121 22%
On-site solar generation 5704 5991 -5%
Net Energy -528 870 -161%

Net energy is the amount generated through solar panels minus the total energy use. So, a negative number means the scenario does not reach zero energy, while a positive number means that the home would generate more energy than it uses.


What Was Learned

While each energy model produced slightly different results, both models reveal a significant fact about residential energy use in this home. The amount of energy consumed by plug loads, lights and appliances dwarfs that of space heating, cooling, and water heating. It’s no surprise because this home includes a highly-insulated and air-sealed thermal envelope as well as a highly-efficient heat pump water heater and energy recovery ventilation system.

In the design of my home, like most zero energy projects, the efficiency of the thermal envelope, water heating, and lighting, was targeted first. So, the next step in improving energy efficiency in a home like mine would be to focus on appliances and plug loads. In fact, research indicates that cable boxes, game consoles, large screens, and other electronics are taking a big bite of the home energy pie each year.

Since selection of appliances and management of plug loads are not generally under the control of designers or builders, one design solution to address this challenge would be to include a home energy management system in zero energy homes.

The results also suggest another approach. With the price of rooftop solar systems declining rapidly, it would be wise to weigh the cost of adding more solar with the cost of other energy saving measures. At some point, adding a few more PV panels may be cheaper than building thicker walls, buying super-efficient windows, or installing energy management systems. Of course, solar electric systems might be limited by more than cost. Solar exposure and available roof area must also be considered. Here again, an energy model can help balance various design issues.


How  the Modeling Programs Compare

In the categories of space heating, water heating, and PV production, the two models used were in fairly close agreement. The methodology for calculating these parameters is well established, so the level of agreement makes sense.

One number that jumps off the comparison table, however,  is the difference in cooling energy. REM/Rate estimated nine times more consumption for cooling than Build Equinox Zero. While the percentage was very high, the absolute number of kilowatt hours was very low due to the almost insignificant cooling requirement at this location.

REM/Rate’s estimate was closer to reality on solar electricity generation by predicting only 5% more kWh than actually was produced and was very close to the 5,767 kWh that the solar contractor estimated for my solar production.

As mentioned, the category of plug loads, lights, and appliances represents by far the largest energy use. REM/Rate predicted 26% more energy used in this category than Build Equinox Zero. This difference may result from the set of assumptions used by each program.

While total energy consumption varied significantly between the two models themselves and between both models and the actuals, virtually all of these discrepancies can be attributed to the lights, electronics, and appliances category. While the percent discrepancies may seem high, both programs offer useful information for design decision making.


Model Evolution

On-site storage batteries and electric vehicles are entering the market in a big way. The next developmental step for energy modeling software should be to integrate these technologies into their calculations. The tight integration of solar, storage, and transportation, along with the usual energy saving measures, will be essential to reach zero energy/zero carbon in the residential sector.

As  momentum is growing to move the housing industry to zero energy for all new construction, building cost-effective zero energy homes is crucial. Energy modeling will be a key step in  achieving this goal. Based on my comparison of actual energy use in my home with these two different modeling systems, the modeling program selected may be less important than the practice of using energy modeling in an iterative manner to optimize performance and affordability of zero energy homes.