Green Star Reference Building HVAC Systems
When you're chasing Green Star certification in Australia, you're not just proving your building uses less energy than average. You're proving it uses less energy than a specific, carefully defined reference building that follows prescriptive rules from ASHRAE 90.1 Appendix G, tweaked for Australian conditions. Understanding these reference building rules matters because they establish the baseline your actual design gets compared against.
The reference building isn't your building. It's a parallel-universe version with the same geometry, same orientation, same occupancy, but with HVAC systems selected according to a rigid formula based on building size, use type, and cooling capacity. You model both buildings, run annual energy simulations on both, and the difference determines your Green Star points. If the reference building rules seem arbitrary in places, that's because they are. But they're consistently arbitrary, which is the point.
How System Type Gets Assigned
Reference building HVAC system type depends on three things: project type, conditioned floor area, and total cooling capacity.
Residential
1
Air-cooled split DX heat pump systems (one per apartment)
Residential common areas
-
Not conditioned
Non-residential and < 2,300 m²<br>Base building systems (Green Star Buildings)
1
Air-cooled packaged reverse cycle heat pump systems<br>No provision for tenant supplementary plant
Non-residential > 2,300 m² and < 1,000 kWr cooling<br>Base building systems (Green Star Buildings)
2
Variable air volume with reheat<br>Air-cooled chillers<br>Tenant supplementary condenser water system (where provided in proposed)
Non-residential and > 1,000 kWr cooling<br>Base building systems (Design & As Built)
3
Variable air volume with reheat<br>Water-cooled chillers<br>Tenant supplementary condenser water system (where provided in proposed)
Fire Stations
1
Package/Split DX reverse cycle (heat pump) systems
For residential projects, you get System Type 1: air-cooled split DX heat pumps, one per apartment. Residential common areas aren't conditioned in the reference model, regardless of what you've actually designed. This sometimes creates interesting conversations when your proposed design includes heated lobbies or conditioned corridors.
For non-residential projects under 2,300 m², you also get System Type 1, but as packaged reverse-cycle heat pump systems rather than splits. No provision for tenant supplementary plant, which can matter if your actual design includes supplementary cooling for server rooms or other high-density loads.
Between 2,300 m² and 1,000 kWr cooling capacity, you jump to System Type 2: variable air volume with reheat, served by air-cooled chillers and a hot water boiler. Above 1,000 kWr cooling capacity, you move to System Type 3, which swaps air-cooled chillers for water-cooled units with cooling towers.
Fire stations get System Type 1 regardless of size, apparently because someone decided packaged heat pumps are the standard practice for that building type.
The Four System Types Defined
1
Package/Split DX reverse cycle (heat pump)
Constant volume
Direct expansion
Electric heat pump + direct electric auxiliary heating
2
Air-cooled chillers
Variable volume
Chilled water
Heating hot water fossil fuel boiler
3
Water-cooled chillers
Variable volume
Chilled water
Heating hot water fossil fuel boiler
4
Package reverse cycle (heat pump)
Constant volume
Direct expansion
Electric heat pump
System Type 1 is packaged or split DX reverse-cycle heat pumps with constant volume fans, direct expansion cooling, and electric heat pump heating with direct electric auxiliary backup. This is your rooftop package unit or split system serving individual zones.
System Type 2 shifts to central plant: variable volume air distribution, chilled water cooling from air-cooled chillers, and hot water heating from a fossil fuel boiler. The fossil fuel requirement is natural gas, regardless of what your actual building uses. This creates some interesting distortions when your proposed design uses something else.
System Type 3 is identical to Type 2 except the chillers are water-cooled instead of air-cooled, which means you're also modelling cooling towers, condenser water pumps, and water treatment. Better efficiency, more complexity, bigger plants.
System Type 4 is packaged reverse-cycle heat pumps with constant volume fans and direct expansion, but without the auxiliary electric heating that Type 1 includes. This system type doesn't appear in the project type assignment table.
The guide notes that reference systems may change as fossil fuel systems get phased out. That's a polite way of saying these rules were written when gas heating was considered normal and might not age well as electrification becomes standard practice.
General Requirements That Apply to Everything
Ten requirements apply regardless of which system type you've been assigned. These establish the baseline configuration and control strategies.
System type and zoning gets specific. For Types 1 and 4, each thermal block gets its own HVAC system. For Types 2 and 3, you can group identical thermal blocks across floors, but Class 5 (office) and Class 9b (education) buildings need one system per perimeter orientation where orientations differ by 45° or more. Everything else gets one system per floor.
Spaces with significantly different loads or schedules (30% load difference or 40 equivalent full-load hours per week schedule difference) need separate systems: single-zone if under 500 m², independent VAV if over 500 m². This prevents the reference building from averaging out high-density server rooms or 24-hour spaces with normal office areas.
Equipment efficiencies come from NCC Section J or MEPS requirements, matched to equipment type and capacity. You're not modelling best-in-class equipment. You're modelling minimum code-compliant equipment, which is deliberately mediocre to make your actual design look better by comparison.
Equipment sizing is based on design loads calculated by the simulation, and you need to report unmet load hours to prove the system can actually condition the space. If your reference building shows thousands of unmet hours, something's wrong with the sizing or the model.
No preheat coils in the reference building, even if your actual design includes them. This is one of several places where the reference building deliberately differs from what you might actually design.
Fan operation follows NCC Section J requirements. Economisers get included where NCC Part J5 requires them. Outside air rates come from NCC Part F4, kept consistent with the baseline used for the indoor air quality credit to avoid double-counting benefits.
Ventilation heat recovery or demand controlled ventilation gets included where NCC Section J requires it. You don't get to omit these just because they're optional in the proposed building.
Supply airflow rates are calculated from an 11 K supply-to-room temperature difference, or minimum ventilation requirements, whichever is greater. This is warmer than many actual designs use (which typically target 13-15 K), so the reference building moves more air than might be strictly necessary.
Fan power calculation is detailed and iterative. You start with design pressure loss without adjustment (unless you want to do the detailed calculation per J5.4(c) and (d), which is allowed but not mandatory). Flow rate comes from the load and the 11 K temperature difference. Fan system efficiency comes from J5.4(b) minimums for the actual fan type and installation.
Then you iterate: start with efficiency of 0.45, calculate motor power, recalculate efficiency, adjust motor power, repeat until successive values differ by less than 0.1, which usually takes two or three iterations. The formula is straightforward once you've done the iteration:
P_ref = (q × Δp) / (1,000,000 × η_system)
Where P_ref is reference pump power (kW), q is duty flow rate (L/s), Δp is system duty pressure loss (Pa), and η_system is system efficiency.
System Type 1: Package/Split DX reverse cycle (heat pump) systems
System Type 1 covers packaged or split DX heat pumps. Each thermal block gets its own system. The configuration is constant volume fan, DX heating and cooling coils, and electric auxiliary heat for backup.
Heat pump control uses multistage space thermostats plus an outdoor air thermostat that only energises auxiliary heat on the last thermostat stage when outdoor temperature drops below 4°C. This prevents the reference building from defaulting to expensive resistance heat during mild weather when the heat pump could handle the load.
Performance uses minimum NCC Section J or MEPS values for EER (cooling) and COP (heating). If only EER is specified, assume COP equals EER. If AEER and ACOP are given, use those values directly.
Part-load curves should match the specified unit if you know what equipment you're using. If you're modelling before equipment selection, use standard part-load curves from the simulation software, but apply the same curves to both proposed and reference models. For reverse-cycle heat pumps where no typical curves exist, you can model them operating uniformly at full-load EER/COP, which is generous but acknowledged as acceptable given data limitations.
Fan configuration is either cycling (on with the coils, off when neither heating nor cooling) or continuous operation. Calculate fan power per the iterative method in Item 10, based on airflow determined by the 11 K temperature difference.
Direct electric heating where applicable gets 100% efficiency because thermodynamics.
Special spaces that differ significantly from the rest of the building need separate systems. Under 500 m², add another System Type 1 unit. Over 500 m², add an independent VAV system following Type 2 or 3 requirements.
System Type 2: Air-Cooled Chillers
System Type 2 is for buildings between 2,300 m² and 1,000 kWr cooling capacity. Central plant with air-cooled chillers, gas boilers, and variable air volume distribution.
Chiller setup is two chillers at 55% of design cooling capacity each. They stage when the first unit hits 100% load. Performance uses minimum NCC Section J Option 2 or MEPS values for EER and IPLV, adjusted for non-standard water temperatures. Part-load multipliers come from the capacity range table: apply these to full-load EER to get performance at 75%, 50%, and 25% load.
Chilled water loop runs at 6.5°C supply and 12.5°C return for design conditions. Operational reset varies by outdoor temperature: below 16°C outdoor you run 10°C supply, between 16°C and 24°C you reset linearly from 10°C down to 6.5°C, above 24°C you run 6.5°C supply. Variable primary flow with minimum 70% through each chiller. Pump power calculated per Item 10, apportioned by chiller capacity.
Boiler setup is two units at 60% of design heating capacity each, burning natural gas regardless of your actual fuel source. Minimum NCC Section J efficiency at full load. Part-load efficiency varies linearly from full load at 100% output down to 70% of full-load efficiency at 15% output.
Hot water loop runs at 80°C supply and 65°C return for design conditions. Operational reset: below 8°C outdoor you run 80°C supply, between 8°C and 16°C you reset linearly from 80°C to 60°C, above 16°C you run 60°C supply. Constant primary flow through each boiler, variable speed pumping overall.
AHU configuration includes cooling coil on chilled water, no preheat coil even if your proposed design has one, economiser if NCC requires it, and supply airflow based on 11 K temperature difference.
VAV boxes have heating coils on hot water, minimum 40% airflow or outdoor air requirement (whichever is greater). Zoning is one system per perimeter orientation for Class 5 and 9b buildings where orientations differ by 45° or more, one system per floor for everything else.
Supply fans are variable speed with Method 1 part-load performance curves from ASHRAE 90.1 Appendix G. This accounts for the fact that fan power doesn't drop linearly with flow because static pressure behaviour isn't linear either.
If your proposed building includes tenant supplementary heat rejection, add a single-cell cooling tower with staged primary/secondary pumping, heat exchanger, and controls targeting 25°C leaving water.
System Type 3: Water-Cooled Chilles
System Type 3 applies when design cooling load exceeds 1,000 kWr. This is System Type 2 but with water-cooled chillers and cooling towers instead of air-cooled units.
Chiller setup is three units: 45%, 45%, and 15% of design cooling capacity. Staging sequence is 15% only, then 45% only, then 45% plus 15%, then both 45% units, then all three. This runs small loads efficiently and only fires everything during peak conditions. Performance uses minimum NCC values for EER and IPLV, with part-load multipliers from the capacity and compressor type table (different curves for screw versus centrifugal compressors).
Chilled water and hot water loops are identical to System Type 2. Same design temperatures, same reset schedules, same pump configurations.
Cooling towers are two axial fan units at 50% of design heat rejection each, with variable-speed fan control. Design leaving temperature is 29.5°C or 5.5 K approach to design wet-bulb temperature, whichever is greater, with 4.5 K range. Control target is more aggressive: maintain 20°C condenser water with minimum 3 K wet-bulb approach. This improves chiller efficiency during most operating hours.
Condenser water loop maintains constant primary flow through each chiller. Pump power gets apportioned by heat rejection duty including any tenant supplementary systems. Tower fan power meets NCC requirements for axial fans.
AHU, VAV, and fan configurations match System Type 2 exactly. Same coils, same zoning rules, same minimum flows, same part-load performance curves.
Tenant supplementary heat rejection follows the same requirements as System Type 2 if your proposed building includes it.
Hot Water System Parameters
For both System Types 2 and 3, hot water operates under these conditions:
< 8°C
80°C
8-16°C
Linear variation from 80°C to 60°C
> 16°C
60°C
Design is 80°C supply and 65°C return, which is fairly standard for non-condensing systems but warmer than you'd run a condensing boiler if you wanted it to actually condense. The reset schedule helps efficiency when outdoor conditions allow it and might let condensing boilers condense during mild weather if return temperatures cooperate.
Pumps use constant primary flow through each boiler with variable speed on the distribution side. Piping losses match whatever you modelled in the proposed building. Thermal inertia scales proportionally with the ratio of peak heating loads between reference and proposed buildings.
Chilled Water System Parameters
For both System Types 2 and 3, chilled water operates under these conditions:
< 16°C
10°C
16-24°C
Linear variation from 10°C to 6.5°C
> 24°C
6.5°C
Design is 6.5°C supply and 12.5°C return, giving a 6 K delta-T which is standard practice. Warmer chilled water when outdoor conditions allow reduces the lift the chiller works against, improving efficiency. The tradeoff is you need slightly more flow to deliver the same cooling, and you might compromise dehumidification during humid conditions.
Pumps use variable primary flow with minimum 70% of design flow through each chiller. This is more aggressive than many real systems run (which often maintain higher minimum flows for chiller protection), but represents good practice for energy efficiency. Pump power gets apportioned in the same ratio as chiller capacities.
Tenant Supplementary Heat Rejection
For System Types 2, 3, and 4 (where provided in the proposed building), tenant supplementary systems follow specific requirements.
Configuration is a dedicated single-cell axial fan cooling tower sized for 100% of duty, with staged primary/secondary variable-speed pumping (two pumps at 50% each on both primary and secondary sides) and a heat exchanger separating the open tower circuit from the closed building circuit. Total pump power splits 1:2 between primary and secondary pumps.
If you don't know actual heat rejection duty, assume 50% of design load Monday to Friday from 8am to 6pm, and 20% of design load at all other times.
Design leaving temperature
29.5°C or 5.5 K approach, whichever is greater
Range
4.5 K
Tower leaving water target
25°C
Minimum wet-bulb approach
3 K
Heat exchanger ΔT
1 K
Secondary system flow temperature
26°C (nominal)
Minimum pump flow (individual)
30% of design
Minimum tower flow
40% of design
The minimum flows prevent damage from running equipment too slowly but limit turndown and part-load efficiency.
Part-Load Performance Tables
The tables define multipliers to convert full-load EER into part-load EER at 75%, 50%, and 25% of capacity. This lets you model IPLV performance using full-load data.
Air-Cooled and Water-Cooled Chillers by Capacity
75%
50%
25%
75%
50%
25%
< 350
1.143
1.556
1.521
1.192
1.368
1.089
350-499
1.143
1.556
1.521
1.053
1.142
1.142
500-699
1.143
1.556
1.521
1.112
1.232
1.232
700-999
1.235
1.736
1.725
1.107
1.174
1.181
1000-1499
1.235
1.736
1.725
1.107
1.174
1.181
≥ 1500
1.235
1.736
1.725
1.053
1.142
1.142
Air-cooled chillers show substantial efficiency improvements at part load (multipliers above 1.5 at 50% load), reflecting that they're working against lower condensing temperatures when they're not running full-out. Water-cooled chillers show smaller but still significant improvements.
Chiller Type and Part Load Multipliers
Heat Rejection
Compressor
Min
Max
75%
50%
25%
Air-cooled
Any
0
528
1.59
1.79
1.2
Air-cooled
Any
> 528
n/a
1.62
1.83
1.22
Water-cooled
Screw
0
264
1.51
1.71
1.16
Water-cooled
Screw
264
528
1.49
1.69
1.11
Water-cooled
Screw
528
1055
1.52
1.68
1.17
Water-cooled
Screw
1055
2110
1.48
1.68
1.14
Water-cooled
Screw
> 2110
n/a
1.49
1.7
1.15
Water-cooled
Centrifugal
0
528
1.54
1.75
1.13
Water-cooled
Centrifugal
528
1055
1.39
1.59
1.12
Water-cooled
Centrifugal
1055
1407
1.49
1.68
1.11
Water-cooled
Centrifugal
> 1407
n/a
1.49
1.7
1.15
Screw and centrifugal compressors have different curves because they unload differently. Screw compressors use slide valves and show relatively linear unloading. Centrifugal machines use inlet guide vanes and show efficiency losses at very low loads but good efficiency in the 25% to 75% range where they spend most of their operating hours.
Why These Rules Exist
The whole point of defining a reference building this prescriptively is to create a consistent baseline that doesn't give credit for things that should be standard practice anyway. If the reference building used best-practice systems, you'd only get Green Star points for truly exceptional design. If it used worst-practice systems, everything would look good by comparison and the rating would be meaningless.
The system types attempt to represent what a competent but uninspired engineer would design if they were only trying to meet minimum code requirements. The efficiencies are code minimums. The controls are basic. The staging is simple. It's not bad design, but it's not particularly clever either.
Where this creates frustration is when your actual building type or climate doesn't fit the assumptions. If you're in Darwin and the reference building specifies a gas boiler for a building that realistically has no heating load, you're comparing against something that wouldn't get built. If you're designing a laboratory with 100% outside air and the reference building assumes recirculation, the comparison gets distorted.
But the rules are the rules, and they're documented thoroughly enough that you can model them consistently. Which is really all a baseline needs to be: clear, repeatable, and equally annoying to everyone.
Last updated
Was this helpful?