Energy Modeling Software Compared for MEP Design Teams

The question owners ask is usually some version of “which energy modeling software should our team be using?” It sounds like a procurement question. It is really four questions in one, because the same phrase covers tools that size ductwork, tools that produce a Title 24 Certificate of Compliance, tools that support a LEED submission and tools that test a massing concept in an afternoon. Buy the wrong one and nobody finds out until plan check.

This article compares the tools that matter for commercial work in California, separates the engines from the interfaces built on them, and sets out which programs are legally usable for which purpose. Budlong runs this analysis on live projects as part of our MEP engineering services, and most of what follows comes from watching which tool choices cost a project time. The wider range of that work sits on our expertise overview.

1. What Is Energy Modeling Software?

Energy modeling software predicts how much energy a building will consume across a full year, before it exists. It divides the building into thermal zones, solves a heat balance on each zone for every hour of a typical weather year, and couples that to models of the HVAC systems, plant, lighting, water heating and plug loads serving those zones. A full annual run covers 8,760 hours.

The category exists because prescriptive checklists cannot handle trade offs. A checklist can say a wall must reach a given insulation value. It cannot say whether a heavily glazed south facade is acceptable when the plant is exceptional and the lighting controls are aggressive. Only a simulation answers that, which is why every modern energy code has a performance path and why that path requires a model.

In practice the software gets asked to do three different jobs, and no tool is equally good at all three.

Regulatory compliance

Producing the documentation a permit or certification requires: a Title 24 Part 6 Certificate of Compliance, an ASHRAE 90.1 Appendix G Performance Cost Index, a federal tax deduction calculation. This work is rule bound, the tool must be permitted for the purpose, and elegance counts for nothing here.

Design decision support

Answering questions the team is still arguing about: does a deeper overhang beat better glass, does radiant with dedicated outdoor air beat variable air volume, is a heat recovery chiller worth the plant room space. Speed matters more than polish.

Operational analysis

Calibrating a model against metered data to justify a retrofit or verify savings. This is where commissioning services and modeling meet, and it is a different discipline.

2. How It Works: The Engine Underneath the Interface

The most useful thing an owner can understand here is that nearly every product is two things stacked together. Underneath is a simulation engine, the code that does the physics. On top is an interface, the software a person uses to describe the building and read results.

The engines

EnergyPlus is the U.S. Department of Energy’s engine, developed with the national laboratories and released as open source. It solves the zone heat balance and the system and plant models simultaneously rather than in sequence, which matters for variable refrigerant flow, radiant systems, dedicated outdoor air and thermal storage. It reads a text input file and writes results, with no interface of its own.

DOE-2 is the older lineage, developed at Lawrence Berkeley National Laboratory and still widely used through its DOE-2.1E and DOE-2.2 branches. It runs loads, then systems, then plant in sequence. That structure is fast and produced much of the industry’s benchmark data, but it handles feedback between zone and system less naturally.

The California Simulation Engine is a purpose built engine used inside the Energy Commission’s compliance software, mainly on the residential and multifamily side, alongside EnergyPlus for nonresidential work.

Why the distinction matters commercially

Two products on the same engine will not necessarily give the same answer for the same building. The interface decides how geometry is translated, which libraries load by default and how autosizing is invoked. A published number reflects those assumptions as much as the engine’s physics, which is why “it uses EnergyPlus” is a statement about pedigree, not accuracy.

Validation is the counterweight. ANSI/ASHRAE Standard 140, maintained by ASHRAE, is the method of test used to check simulation programs against benchmarks, and reputable vendors publish their results against it.

3. The Tool Landscape, Compared

The tools group into four families, and the family predicts behaviour better than any feature list.

General purpose simulation interfaces

OpenStudio is a software development kit and application layer for EnergyPlus and the Radiance daylighting engine, originally from the National Renewable Energy Laboratory. Its strength is scripting and automation, which makes it the natural choice for parametric studies and repeatable workflows, and it expects a user comfortable with structure.

eQUEST is the long standing free graphical interface to DOE-2, with wizards that build a model quickly from a handful of inputs. It stays popular for utility incentive work and fast comparative studies, and its limits show on systems DOE-2 was never designed for.

IES Virtual Environment is a commercial suite built on the vendor’s own ApacheSim dynamic thermal engine, with daylighting, computational fluid dynamics, egress and solar modules in one environment. It is the most expensive mainstream option and the most capable across unusual physics. DesignBuilder is a graphical interface to EnergyPlus with Radiance daylighting and CFD, and the friendliest route in for a team without a scripting habit.

Manufacturer design tools

Trane TRACE 3D Plus replaced the older TRACE 700 engine with EnergyPlus and follows the sequence a design engineer works in: load design, system selection, then energy and economics. Carrier HAP uses Carrier’s own hourly calculation with ASHRAE load methods and is validated against ASHRAE Standard 140. Both are strong at sizing and system comparison. Neither is a compliance tool in California.

Compliance front ends

CBECC is the Energy Commission’s own public domain compliance software, connecting to both the California Simulation Engine and EnergyPlus. EnergyPro from EnergySoft wraps the same rules in a modular commercial product, using EnergyPlus for its Title 24 performance module and a licensed DOE-2.1E engine for ASHRAE 90.1 work. Right-Energy Title 24 serves single family residential, and IES VE Title 24 is the compliance configuration of the IES suite.

4. Which Projects Need Which Kind of Model

Two distinctions decide the answer, and both are routinely collapsed by buyers.

Load calculation is not energy modeling

A load calculation sizes equipment. It solves a small number of extreme design conditions to produce airflow in cfm, cooling in tons, heating in Btuh and pump flow in gpm, room by room. Its outputs feed the equipment schedule, the duct layout and the service size.

A whole building energy model predicts consumption. It solves an entire weather year with simplified zoning and averaged schedules to produce annual kilowatt hours, therms, cost and carbon, feeding a compliance report or a design decision.

The failure modes are specific. Size equipment from an annual model and you inherit smeared zoning, generic schedules and autosized results carrying none of the room by room air distribution logic a design needs. The common version is one model asked to do both, zoned coarsely for compliance and therefore useless for duct design. Our note on HVAC load calculation methods covers the sizing side.

Early design tools versus detailed design tools

Early stage modeling works on massing: simple boxes, a handful of zones, orientation and window to wall ratio swept parametrically. Runs take minutes and the tolerance is wide, but the value is large because the decisions with the biggest effect on annual energy are architectural and get made before any equipment is chosen.

Detailed modeling works on the real building: real zoning, real system types, real control sequences, real plant. Runs take longer, the tolerance is narrow, and this is the model that goes into a compliance submission.

The mistake is carrying an early massing model into a regulated submission because it exists and the fee is tight. Early models rest on assumptions no reviewer will accept, and retrofitting them costs more than rebuilding. Plan for two, and run the analysis alongside HVAC design services rather than after them.

5. What a Model Asks of Each Building System

A model is only as good as what it is fed. Where one is wrong, one of these inputs was assumed rather than confirmed.

Architecture and envelope

Geometry, thermal zoning, construction assemblies with real layer properties, window to wall ratio by orientation, glazing U factor and solar heat gain coefficient, external shading, thermal mass and infiltration. Infiltration is the input most often left at a default, and it moves heating energy more than expected.

Mechanical

System type and configuration, terminal unit type, plant equipment with part load performance curves rather than single point efficiencies, economizer strategy, demand control ventilation, heat recovery effectiveness, and the control sequences themselves. Standardised sequences such as ASHRAE Guideline 36 make this far more reliable, because modeler and controls contractor then describe the same thing.

Electrical and lighting

Lighting power density by space type, control credits and their triggers, receptacle and plug load densities, elevator loads and process equipment. Plug loads are often the second largest end use in a modern office and are almost always modeled from a table rather than measured. That connects to electrical engineering services and the load schedule behind the service size.

Plumbing and service water heating

Hot water demand profile, distribution and recirculation losses, storage volume and standby loss, and equipment type. Heat pump water heating changes the energy result, the space requirement and the electrical service, so settle it early.

Solar photovoltaics and storage

Array area, tilt, azimuth, module efficiency, inverter losses, shading from parapets and rooftop equipment, and battery dispatch. Under the 2025 California code these are code driven inputs for many nonresidential occupancies, which puts photovoltaic design inside the model from schematic design onward.

Controls and operating schedules

Occupancy, lighting, equipment, thermostat setpoints, setback strategy, optimal start and unoccupied mode behaviour. Schedules are the largest single source of variance between a model and a metered building, and the input with the least documentation behind it.

6. The California Constraint and ASHRAE 90.1 Appendix G

This section decides the shortlist for anyone building in California, and it is the part most commonly discovered too late.

Title 24 Part 6 performance path

Performance path compliance cannot be demonstrated with whatever energy modeling software the team happens to own. The California Energy Commission approves specific programs, at specific version numbers, as alternative calculation methods. Anything else produces a document a plan checker cannot accept.

Permit applications submitted on or after January 1, 2026 must use software approved for the 2025 Energy Code. The Energy Commission’s approved software list is authoritative and is updated as versions come and go.

Three consequences follow. EnergyPlus, OpenStudio, eQUEST, DesignBuilder, TRACE 3D Plus and Carrier HAP all sit outside the compliance path in California. Approvals are version specific and can expire, so “we have CBECC” is not the same as “we have an approved build”. And the scope limits bite: a nonresidential alteration cannot use the IES VE Title 24 approval.

ASHRAE 90.1 Appendix G for LEED

Appendix G works on a different principle. There is no approved list. Section G2.2 of ASHRAE Standard 90.1 instead sets capabilities the program must have: an 8,760 hour simulation, hourly variation in occupancy, lighting, equipment power and setpoints, thermal mass, sufficient zones, part load curves, capacity limits with unmet load hours reported, and testing to ASHRAE Standard 140. Any program meeting them can be used.

The baseline differs too. Where the Title 24 baseline is a California specific building rebuilt at the prescriptive minimum, the Appendix G baseline is stable, does not move with each code revision, and gives credit for exceeding standard practice the code does not regulate. The result is a Performance Cost Index measured against a target, and the DOE Building Energy Codes Program documents the difference.

LEED versions layer on top. LEED v4.1 references 90.1-2016. LEED v5 moves to 90.1-2019 or 90.1-2022 and shifts the headline metric from pure cost toward a source energy basis, which changes how electrification scores. The U.S. Green Building Council publishes the current requirements.

For a California project pursuing LEED, that means two models. Build both from one shared geometry and one set of system assumptions so the inputs cannot drift apart, and scope the effort once rather than twice. That coordination is the core of our LEED MEP services and the wider sustainable design services around them.

Secondary uses of the same model

Once a compliance grade model exists it supports federal 179D tax deduction analysis, utility incentive applications, CALGreen documentation and the operating cost case behind a net zero building strategy, cheaply if the model was built with them in mind.

7. Healthcare, Laboratories, Data Centers and Aviation

Four building types stress the tools in ways a standard office does not.

Healthcare

Hospitals run continuously and carry ventilation and pressure requirements set for infection control rather than energy. The model must treat air change rates and pressure cascades as fixed constraints, which removes most of the ventilation trade offs a performance path model would exploit. Tools with strong control sequence modelling earn their licence here, as our healthcare facilities work shows.

Laboratories

The 2025 California Energy Code removed the laboratory exemption earlier cycles carried, so lab spaces now sit inside the compliance envelope. High exhaust volumes, once through air and large process loads are where engines diverge and generic templates go badly wrong. Confirm the tool models the exhaust strategy explicitly rather than as a fixed airflow, as we do on our laboratory and industrial projects.

Data centers

Computer rooms are a covered process with their own efficiency and economizer requirements. The IT load profile dominates everything else, rack densities are rising, and liquid cooling is arriving faster than standard component libraries have kept up. Expect custom equipment curves and a careful conversation with the authority having jurisdiction. Context sits under mission critical.

Aviation, transit and education

Terminals and stations bring large volumes, stratification, high door infiltration and displacement ventilation, which is where computational fluid dynamics in the same environment as the energy model earns its keep. Schools meet the multi zone heat pump baseline in the 2025 code head on, making early comparative modelling valuable for districts holding a fixed budget, as our education practice sees constantly.

8. Matching the Tool to the Project

The table below is the shortest honest answer to “which one should we use” in a California context. Where two tools are named, the first is primary and the second covers a scope the first cannot.

Cost and learning curve

Licence cost is the least useful number in this decision. EnergyPlus, OpenStudio, CBECC and eQUEST are free. DesignBuilder sits in the low thousands per seat. EnergyPro is modular, with the complete nonresidential Title 24 package listed by EnergySoft at around $2,400. TRACE 3D Plus and Carrier HAP sit in the middle and often arrive through manufacturer channels, and IES Virtual Environment is the most expensive mainstream option.

Analyst time dominates the total. Every one of these tools takes weeks to reach competence and far longer to reach the judgement needed to defend a result under review. A free tool that takes half again as long is not cheaper, and a firm running one licence and one modeler has a single point of failure on every deadline. Compare cost per finished deliverable, the discipline we apply to scoping MEP drafting services against output rather than seat count.

9. Workflow, Interoperability and Calibration

The tool matters less than the workflow around it. Most modelling problems on real projects are handoff problems, not physics.

The workflow, stage by stage

At concept, a coarse massing model tests orientation, form and glazing ratio. At schematic design, the compliance path and tool are chosen and a first pass model confirms the strategy is viable. At design development, the detailed model is built with real zoning, systems and sequences and starts informing equipment selection. At construction documents it is finalised and the compliance documentation produced. After occupancy it can be calibrated and become an operational asset rather than a permit artefact.

gbXML, IFC and the reality of importing a Revit model

gbXML is the schema most energy tools use to receive geometry from BIM. It carries spaces, surfaces, adjacencies, openings and construction references. It does not reliably carry HVAC system definitions, plant configurations, control sequences or operating schedules. IFC is a richer data model but a weaker practical handoff for energy analysis.

The Revit reality is that a model built for documentation is not a model built for energy analysis. Rooms and spaces must exist and have volumes, and room bounding behaviour must be set on every element that should enclose a space. Curtain wall and sloped glazing often arrive as the wrong surface type, sliver surfaces appear at floor and roof junctions, and anything modelled as a generic mass drops out entirely. A clean import is the result of deliberate BIM setup, not of the exporter.

The rule that survives contact with real projects is simple: import the geometry, rebuild the systems. Round tripping a mechanical system through gbXML costs more in repair than it saves. The discipline that makes geometry work is the one that makes coordination work, which is why this sits close to advancements in BIM for MEP coordination and clash detection practice, and runs on into 5D BIM and cost management.

Calibration and measurement tooling

A model of an existing building is a hypothesis until it matches metered data. ASHRAE Guideline 14 sets the accepted criteria: with monthly data, normalized mean bias error within 5 percent and coefficient of variation of the root mean square error within 15 percent; with hourly data, 10 percent and 30 percent. Hitting those numbers needs interval data, a trend export and enough submetering to separate end uses.

This is where modelling and commissioning converge. The trend data supporting calibration is the data supporting functional performance testing, and the sequences a calibrated model needs are the ones a commissioning agent already verifies. Running both as one scope produces a better model, a cheaper commissioning exercise and the baseline that makes a later retrofit case defensible. That is the argument for building sustainability analysis into delivery rather than bolting it on.

Key Takeaways

• Energy modeling software is two layers. Choose the interface for the deliverable you need, then check what engine sits underneath it.
• Only CEC approved software can produce a Title 24 Part 6 performance path submission. For the 2025 code that means CBECC, EnergyPro 10, IES VE Title 24 2025 or Right-Energy Title 24.
• Approvals are version specific and carry scope limits. IES VE Title 24 covers nonresidential new construction only; Right-Energy covers single family new construction only.
• ASHRAE 90.1 Appendix G has no approved list, only the capability requirements in Section G2.2 and validation to ASHRAE Standard 140.
• A Title 24 model does not satisfy LEED and an Appendix G model does not produce an NRCC. Plan for two models built from shared inputs.
• Load calculation tools size equipment; whole building simulation tools predict consumption. Using one for the other’s job fails predictably.
• Import geometry through gbXML, rebuild systems natively. Systems, plant and sequences do not survive the handoff.
• Licence cost is a small share of total cost. Analyst time and review defensibility dominate, and a single modeler is a single point of failure.

10. Where Budlong Works

Modelling requirements are set statewide, but the review culture around them is local. Plan check expectations differ between jurisdictions, local reach codes add requirements above the state baseline, and utility incentive programs vary by service territory. Budlong maintains teams across California so local review experience sits behind every submission, as our company story page describes.

Sectors where the tool choice matters most

Healthcare constrains the ventilation trade offs a model would normally use. Laboratories lost their exemption in the 2025 code and now need explicit exhaust modelling. Data centers push equipment libraries past their design intent. Schools meet the multi zone heat pump baseline directly.

11. Related Reading

12. Frequently Asked Questions