Aris Talos

• Multiple HPXML rating requests. The rating engine can evaluate each request declared in one HPXML file, including home energy ratings, CO₂ index, ENERGY STAR, IECC ERI and DOE Efficient New Homes. Results retain the requested program and rules edition and identify requests that cannot complete.
• Program results include the comparison home. ENERGY STAR and DOE Efficient New Homes calculations evaluate the Rated Home and the program Target Home, apply the required adjustments and report a program verdict.
• Faster calculations for a detailed home. Talos prepares solar shading and annual room and foundation calculations more efficiently. A measured large-home run is faster; times vary by home and machine.
• Rain follows recorded precipitation. Talos wets exposed surfaces from recorded precipitation depth, avoiding lingering weather codes that marked dry, sunny hours as wet and distorted heating and cooling estimates.

• Faster startup and annual runs. Talos uses less memory at startup and avoids repeating foundation, window and equipment calculations during annual runs. Equipment-only edits can reuse the building model, reducing the work needed to update an energy rating.
• Broader heating and cooling support. Talos can rate more homes with multiple HVAC systems, mixed heating fuels, boilers, fan coils, electric resistance heat, and separate heating and cooling equipment. Calculations preserve each system’s equipment and distribution inputs, including independent fan speeds, capacities and cycling behavior.
• Expanded solar and hot-water calculations. Supported HPXML configurations now include detailed solar water heating, photovoltaic generation, batteries, and water heating that recovers heat from air-source or ground-source heat pumps. Heat-pump water-heater tank controls now respond when backup elements switch on or off. Dehumidifier electricity and moisture removal also enter the annual calculation.
• Apartment-level building calculations. The building model now reports heating, cooling and hot-water demand for individual apartments and common spaces, with operating assumptions that can vary by dwelling.
• More faithful construction and equipment modeling. Corrections cover foundation heat transfer, wind exposure, roof boundaries, shared surfaces, windows, skylights, shading, lighting heat gains, ventilation and duct leakage. Reference homes preserve prescribed construction and opening areas across irregular room shapes; rated homes apply shading to each window and use consistent ventilation flow for heat and humidity.
• A more useful HPXML runner. Uploaded homes use their authored equipment for supported rating configurations. Results show equipment energy use and hourly playback, while identifying the weather used for the rating separately from the interactive building view. Exported HPXML files now name that rating weather file.
• Expanded validation evidence. The refreshed comparison completed all 421 cases: 64 Building Science Institute cases, 24 EPA test cases and 333 OpenStudio-ERI cases. Annual-load plots distinguish RESNET’s published intervals from Building Science Institute’s ±5% OpenStudio-ERI target, a useful higher bar championed by Building Science Institute. Across the 333 OpenStudio-ERI cases, median absolute ERI difference is 0.115 points, the 90th percentile is 0.352 and the maximum is 0.836. All 26 annual-load comparisons are inside both bands, and all 33 hot-water criteria pass. Criteria outside their bands remain visible in the report.

• A wider range of homes can be rated. More HPXML files now complete an ERI calculation, including homes that rely on prescribed appliance and ventilation defaults, derived duct areas, stated duct efficiencies, or a mix of fixed and automatically sized heating and cooling equipment. Talos applies the defaults supplied by the rating rules where those inputs are omitted.
• Closer agreement with OpenStudio-ERI. On the same 12 homes in our published comparison, median absolute ERI difference fell from 0.24 to 0.11 points, and the largest difference from 0.53 to 0.21. Corrections cover fan runtime, duct heat transfer, equipment staging, reference-home construction, and thermal behavior. We also compare loads and energy use component by component, because a close score can hide errors that cancel out.
• A more detailed 3D view. The interactive house view now includes richer materials, more detailed roofs and equipment, and lighting that follows the sun and moon. Day and night views make it easier to explore the home alongside its energy results.
• Heating sized for an occupied home. Heating design loads now credit a small amount of heat from refrigeration and one sleeping occupant. We accept that an empty home may occasionally dip below its setpoint. The report explains this choice beside the sizing comparison; energy ratings retain their prescribed operating assumptions.
• A clearer energy ratings chapter. The report now explains ERI as a comparison with a standardized reference home and identifies OpenStudio-ERI as the benchmark for our scores.

• Faster energy calculations and ratings. Faster ground-temperature initialization reduces preparation time for first-time calculations. Talos also runs more work concurrently across browser workers and reuses unchanged calculations after edits, including solar calculations when only insulation changes. In a production Chrome test on one home, changing wall insulation and updating the energy report and rating took 1.77 seconds, down from 2.57 seconds, a 31% reduction averaged over two runs before and after. Results vary by home and device.
• Better conditioned-crawlspace modeling. Conditioned crawlspaces now contribute to the home’s enclosed air volume and leakage calculations without being treated as occupied rooms. Their temperatures respond to the surrounding construction and ground.
• Heat-pump water heaters coupled to the home. Supported systems now calculate tank operation alongside the building’s hourly conditions, including heat and moisture removed from indoor air and the resulting effect on heating and cooling.
• More representative Washington energy-code comparisons. For supported air-to-water heat-pump designs, WSEC R405 calculations now evaluate the proposed hydronic system directly instead of substituting an air-source system. Results show its compressor, pump, fan, backup-heating and hot-water energy use, alongside unmet loads and calculation assumptions.
• Radiant-aware airflow sizing. Talos calculates room airflow requirements with the contribution of radiant heating and cooling included, helping distinguish the load served by water from the load remaining for air.
• More complete hydronic calculations. Distribution-module calculations now account for coupled pumps and circuits, with continuous pump and coil performance curves.

• Broader HPXML support. Talos now runs more HPXML configurations through room simulation, equipment sizing and energy ratings. These include apartments, multistory homes, furnaces and air conditioners, central and ductless heat pumps, and equipment with fixed or automatically determined capacity. It preserves the construction and system inputs from each file and reports configurations that are not yet supported.
• More inputs in the annual calculation. The annual model now includes more detail for foundations and crawlspaces, roofs and shading, hot water, HVAC controls, ventilation, solar, appliances and ceiling fans.
• Duct and hydronic network calculations. Talos calculates the air delivered to each vent and the water delivered to each radiator or floor loop using the complete network and the selected fan or pump.
• Faster recalculation after edits. Annual simulations and repeated ratings reuse results for unchanged inputs, reducing recomputation after each edit.

The first public release of the engine, and of the evidence behind it. At launch Talos is judged against three standards: ANSI/ASHRAE Standard 140, the envelope and fabric load tests; ACCA Manual J, for whether the sizing procedure is followed as written; and RESNET Publication 002-2024, the accreditation suite home energy ratings are issued under. The report publishes what came back from each, including where we do not agree yet.