Refrigerants: Ammonia vs CO2
This article was corrected on February 26, 2013
This article was corrected on February 26, 2013
Arenas are to be seen in almost every municipality across Canada, providing a place for people to enjoy playing or watching hockey, and for other sports and special events. The Canadian Recreation Facilities Council conducted a national census1 in 2005 showing that there are more than 2,450 arenas and 1,300 curling rinks in Canada. The largest construction boom of arenas in Canada occurred throughout the 1970s, and most of them are municipally owned.
The census also found that 65% of ice arenas use ammonia as a refrigerant, 25% use hydrochlorofluorocarbons (HCFCs) (mainly R22), and 10% have natural ice.
HCFSs are ozone-depleting substances and, under the terms of the Montreal Protocol, the Canadian government has adopted a phase-out schedule. Since 2010, no new R22 equipment is manufactured in Canada or imported.
The average amount of synthetic refrigerant released into the atmosphere every year is 60 kg2. An arena using R22 refrigerant generates greenhouse gas emissions (GHG) of over 145 CO2 equivalent tonnes per year.
Due to global environmental concerns and the phase out of HCFCs, engineers are becoming more interested in the use of natural refrigerants in arenas. Safety concerns regarding refrigerants are also an important concern as arenas are often located in residential areas.
CO2 is a natural refrigerant that has recently been considered as one of the most promising for its characteristics, as presented in the table below. Therefore Dessau has recently conducted research of different arenas comparing ammonia and CO2 in their refrigeration systems. This paper outlines our findings.
R-22 and R-717 are widely used as primary refrigerants to cool a secondary fluid, traditionally brine. A tubing network installed in the concrete slab is usually made of 25-mm OD polyethylene pipes. An efficient set-up with secondary fluid, as described in ASHRAE’S 2009 publication “Improving Energy Efficiency in Ice Hockey Arenas”3, is to modify the conventional two-pass arrangement to a four-pass arrangement, therefore reducing the energy consumption of the brine pump (15 HP variable-speed pump instead of 25 HP pump) and consequently reducing the refrigeration load.
We can now use CO2 as the primary and secondary fluid, consequently eliminating the need for a heat exchanger and improving the overall efficiency of the refrigeration cycle. Carbon dioxide can be pumped directly from a low pressure receiver directly into a tubing network in the concrete slab made of 13 mm OD plastic-coated copper tube. A standard 80-ton installation would require a nominal flow rate of 30 GPM, that can be pumped by a 5 HP variable-speed pump. This setup comes with a huge amount of CO2 that needs to be pumped to the receiver and cooled during the off season.
CO2 can also be used only as a primary fluid combined with a conventional secondary fluid. This configuration would be equal to R-717 in terms of efficiency on the evaporating side, yet would significantly reduce the amount of CO2 that needs to be managed during off season.
Cooling towers or evaporative condensers offer the best efficiency with ammonia systems. These pieces of equipment require water treatment and can be a challenge to install on existing roofs due to their heavy weight. The energy recovery concept will define the operating pressures and temperatures.
One interesting characteristic of CO2 is its very low critical temperature (31° C). It opens possibilities that have been studied for years by scientists, such as a transcritical CO2 cycle for refrigeration. Quebec supermarkets have been pioneers in North America of this type of system, which was first installed in Europe. In the last three years, manufacturers offered the technology to arenas, and a small number of systems using a transcritical CO2 cycle have been installed in arenas in Quebec.
Transcritical conditions are reached when operating pressures (and therefore temperatures) are above the critical point for a given fluid. In a transcritical CO2 application, the system won’t have a condenser but has a gas cooler to cool down the CO2. Therefore the global efficiency of the system will strongly depend on how efficiently the heat exchange occurs inside the gas cooler. The lower the gas cooler’s outlet temperature, the higher the compressor’s cooling capacity and efficiency. A COP of 3 can be obtained in winter with a 5° C gas cooler outlet temperature. The efficiency will drop below COP of 2 with an outlet temperature above 30° C, with a year-round average of COP of 2.5. Gas coolers require low maintenance and can easily be fitted on existing structures to replace R-22 condensers.
In most small municipalities, arenas represent the public buildings with the highest annual energy use and consumption — at the same level as public lighting or transport. Around 40% of the total energy consumption goes for heating, mainly using natural gas. Electricity is the second most common energy source, while a few arenas are still using fuel.
With both ammonia and CO2 refrigerants, domestic hot water (showers and ice resurfacer) heating loads to 60°C can be fulfilled with heat recovery by using a high pressure double-wall heat exchanger for CO2 and a desuperheater for ammonia. Other building energy efficiency measures such as low-e ceilings (e.g. installing a thin layer of low-emissivity metallic material just below the structure truss above the ice rink), efficient lighting systems, exhaust heat recovery, and subfloor and snowmelt heating by heat recovery can be applied regardless of the refrigerant chosen.
Being considered as a toxic refrigerant, ammonia can’t be used in a heat recovery condensing coil in a heating system. Therefore, a heat exchanger is required to transfer recoverable energy to a heating loop. Three options can be considered: (1) a tempered 21°C heating loop with water-to-air heat pumps, (2) a direct recovery low temperature (35°C) heating loop with fan-coils, or (3) radiant heating flooring, mainly for stands. Option 1 will benefit the ammonia cycle efficiency by reaching COP of 4 with floating head pressure, despite requiring electric input for the heat pumps. Option 2 will be the most energy efficient option in the context of an existing building, reaching COP of 2.9 year round with a steady condensing temperature of 38°C. Option 3 can be considered as the best strategy in terms of energy for a new arena. Generally, the low water heating temperatures will require larger equipment.
CO2 heat recovery systems benefit from very high temperature gas that can be used in heat recovery coils to heat the spectator stands, perhaps requiring stainless steel piping and precise welding to resist high pressure. The Canadian Standards Association B52 Refrigeration Code states that heat recovery coils should only be installed in large areas such as the stands. As CO2 gas can carry more energy in a given volume, piping size is reduced and there is no need for insulation unless for safety concerns. The installation costs therefore compete with standard heating piping. CO2 certainly offers a little more recoverable energy and at a higher temperature, but the variable heating loads of each arena need to be studied closely to see if such energy is really required on an annual basis.
The chart on page 27 shows the energy consumption for a standard retrofitted 3,250-m2 arena in Montreal The NHL size ice rink has seating for 500 people and operates eight months per year. Heating to 16°C is provided in the stands, while the main hall, players’ rooms, kitchen area and service rooms are maintained at 22°C.
The “Existing arena” in the chart matches the energy consumption of the reference model described in the 2009 ASHRAE publication mentioned above (
The “Efficient arena” includes energy efficiency measures proposed in the same publication. The “CO2 efficient” arena has been modeled with CO2 as the primary fluid plus full energy recovery in heat recovery coils. The “Ammonia efficient” arena has been modeled using a direct recovery low temperature (35°C) heating loop with fan-coils (option 2 as described above in “Energy Recovery”).
How to decide?
In our cold climate transcritical systems using CO2 compete well with ammonia systems in terms of the overall energy consumption, despite the drawback of having a high working pressure. The colder the climate and the more low temperature heat recovery applications (e.g. subfloor heating and snowmelting), the better the performance with CO2 systems.
Ammonia will benefit from any application having high refrigeration loads and low heating loads, as the efficiency of the cycle is much better. For each arena, the energy consumption of ammonia’s pumping and vent fans need to be compared to CO2’s loss of refrigeration cycle efficiency.
In sum, arenas appear to be the turning point where both technologies offer comparable overall performances.
The refrigerant chosen, therefore, will strongly depend on existing conditions such as the space in the mechanical room, pipelines, existing ducts, the structure of the roof, architectural features, and the qualifications of the operating staff. The good thing about both technologies offering comparable performances is that the existing conditions will define which option will better suit a given project and which will allow the municipalities to reduce their installation or operating costs. cce
Mathieu Courchesne, ing. is a mechanical and energy efficiency engineer with Dessau in Longueuil, Quebec.
1 National Arena Census May 2005 – December 2005, Canadian Recreation Facilities Council in association with Hockey Canada and CANMET Energy.
2 Potentiel d’économies d’énergie en refrigeration dans les arenas du Québec [Energy-saving potential in refrigeration in the arenas of Quebec], CEDRL (now CanmetENERGY, Varennes Research Centre).
3 Nichols L. 2009. “Improving energy efficiency in ice hockey arenas.” ASHRAE Journal June 2009.
SIDEBAR, PAGE 25
College Jean-de-Brebeuf’s Arena (3,070-m2) in Montreal was built in 1974 and operates year-round using ammonia as a primary refrigerant. In a recent retrofit, the compressors were replaced with a new refrigeration package, still using ammonia and the existing evaporative condenser. The retrofit involved adding: a direct recovery low temperature (35° C) heating loop using existing pipelines, with new fancoils to heat the stands and other rooms; a desuperheater (heat recovery) for the resurfacer and domestic hot water (arena and adjacent residence for 200 students); an enthalpy wheel for heat recovery of exhaust air to preheat fresh air, and conversion of a 2-pass to 4-pass arrangement for secondary fluid. The arena’s energy consumption improved from 2,993,279 kWh-eq to 1,147,591 kWh-eq. Project costs were $1,365,000. Annual energy savings: $103,805.
SIDEBAR, PAGE 29
Advantages and Disadvantages of
Ammonia vs. CO2
of the refrigeration cycle
Easy to operate, easy maintenance
Well known refrigerant in the industry
Low operating pressure not
requiring special piping
Natural refrigerant, 0 ODP, 0 GWP
Requires a class T mechanical room
Requires special training
for an emergency
Low condensing temperature that
produces low temperature water
for heat recovery
Toxic (heavily regulated in
Heat recovery ammonia coils
can’t be used in systems
Requires a water cooling tower (heavy, water treatment, high maintenance)
No need for class T mechanical room
High temperature fluids
for heat recovery
Heat recovery CO2 coils can
be used for the stands
Gas cooler (lightweight, no water treatment, low maintenance)
Natural refrigerant, 0 ODP, 1 GWP
Medium efficiency of the refrigeration cycle (mainly in summer)
Not very well known refrigerant
in arena industry
High operating pressure requiring stainless steel piping and good welding