Rotary Heat Exchangers Pty Ltd

Technical Aspects


A rotary heat exchanger (rhe) sometimes described as a heat wheel is a rotating wheel, that very efficiently transfers heat from one air stream to another. It is a highly efficient way of transferring thermal energy from a hot to a cold air or gas stream.
Heat is transferred by rotating a porous matrix rotor between the streams while the matrix transfers the heat by conduction and convection into and from the large surface of the matrix.

Note that this can be used summer or winter to recycle vital energy from the exhaust air of a building to air condition the fresh air at very little cost. This is sometimes described as free heating or cooling.

The rhe has efficiency advantages over conventional stationary heat exchangers, such as fixed plate type for example. There is a world wide trend to use rhe's to air condition environmentally sustainable buildings.

A major advantage, which contributes to its high thermal efficiency is that rhe's are 100% counterflow in flow orientation.


Unfortunately the rhe is not well understood because of perceived novelty and more complex nature. I find many consultants are not confident applying the technology.

There are several traps that I would like to explain in simple terms so that all practitioners and contractors in the industry can have a better understanding of the issues involved.


Firstly let me state my position and background in this matter. I am the owner and director of the Australian rhe manufacturing company Rotary Heat Exchangers Pty Ltd (RHE).
I am a professional mechanical engineer with many years of experience in running my own engineering consulting company Ecopower Pty Ltd as well as having worked for CSIRO for 13 years in the research and development of the Australian unique rhe. I therefore have an extensive knowledge of the technology and we are the only manufacturer of rhe;s in Australia. I wear two hats.


There are only two types of rhe available; the  aluminium matrix type which are manufactured in many countries including Japan, USA, China and India; and the Mylar film matrix Australian rhe developed by CSIRO initially and further improved by RHE,

The Mylar wheels have been installed in Australia since 1968 and have a proven long life durability even in aggressive environments such as aquatic centres where they have, in several instances, outlived the 30 year life of the building and were reinstalled in the redeveloped aquatic centre.


Despite the thermal advantage of the counterflow rotating rhe, it has one inherent disadvantage over the fixed type. The rhe has the disadvantage of mixing a small portion of the two flows which is stored inside the internal rotor volume. At any time during rotation each stream is stored in half the volume of the open porous matrix. This type of mixing is for obvious reasons described as carryover and its magnitude is in direct proportion to rotational speed and rotor air space volume.

For this reason the CSIRO Australian Mylar wheel has been designed to provide negligible carryover by making all rotors "thin" thus limiting rotor volume and by maintaining a low rotational speed at optimum performance. All Australian rhe rotors are only 10 cm deep and operate at a low 18 rpm.

In contrast all imported Al matrix rhe's have "fat" rotors varying in depth from 20 to 40 cm with the rotor depth increasing with increasing rotor diameter. Consequently Carryover in these fat rotors may be 2 to 4 times higher than thin Australian heat wheels. The rotor matrix porosities of all rhe are generally similar in order to maximise thermal performance.


The efficiency of heat transfer for rhe's is not so easy to measure. Efficiency is often simply calculated by measuring temperature differences i.e how much the fresh air is heated or cooled as a ratio of the difference between the two approaching fresh and exhaust streams,
Note this only works if the heat exchanger is in balanced flow i.e. fresh and exhaust airflow rates are equal.

There is a complication however. The partial mixing taking place concurrently with heat transfer due to carryover blurs the results taken by temperature measurement. The measured result is a combination of thermal transfer efficiency and mixing transfer and it is difficult to separate these components.

I have published a technical paper on the effect of bypass flows on efficiency in the AIRAH journal, which examines the effects of all by pass flows including carryover on rhe efficiency. This paper gives a theoretical method for correcting the measured efficiency for the effect of carryover.

The publication can be down loaded by clicking on the down load link at the end of this article. In my experience the apparent or measured temperature efficiency of deep rotors can give misleading higher efficiencies of up to 10%, making efficiency correction imperative.


"Fat" rotors are double to 4 times or greater in thickness which can exhibit significant carryover mixing of exhaust air with the incoming fresh air stream. Fat rotors can use a system called purge to reduce this carryover mixing.

Purge is where some fresh air is used to flush out a small segment of the rotor before it crosses into the fresh air side. This is a continuous process and robs the segment of the rotor in the exhaust side of some of the temperature benefit it had acquired thus reducing the overall efficiency.

Furthermore the correct purge flow requires the correct airflow pressure difference to drive this flow and may require the inclusion of an additional fan. For this reason it is seldom used and the true efficiency of the rotor is compromised.


The correct way of specifying a required performance of a rhe, apart from the complexity of carryover, is to specify the required temperature or heat transfer efficiency and pressure drop required by the rhe, given the actual operating conditions. This includes the air flow rates on both sides of the rhe. It is not dissimilar to specifying the performance of a fan, taking an analogy most familiar to the industry.

A fan characteristic exactly specifies the performance duty of a fan and is simply a pressure versus flow curve or table. A heat exchanger or rhe will also have a similar performance characteristic curve or table defining its performance but it will include the third parameter efficiency. Such characteristic must also include the proviso that the rhe is in balanced flow and flow orientation must also be in counterflow i.e exhaust and fresh air streams must flow in opposing direction. This double air flow complexity over a simple fan analogy does pose some interesting questions.

An unbalanced flow heat exchanger (irrespective of heat exchanger type) can exhibit a much higher efficiency than when the flows is balanced. You can increase the efficiency of any heat exchanger simply by reducing the flow on one side. The reason is that by definition the efficiency is based on what happens to the temperature of the low flow side. If we increase the high flow side flow than there is more heat capacity available to impose heat transfer on the low flow side resulting in a higher temperature change on this side. For this reason it is critical to specify all flow conditions and flow orientation when specifying efficiency. To simply specify minimum and maximum performance in isolation may be very misleading.
For example It would be equivalent to stating that a fan has the capability of producing 100l/s to 1000l/s without specifying at what pressure delivery.

Consultants specifying rhe’s should enquire about these details and corrected performance values for their individual design proposals from their preferred manufacturer.

We at RHE have always since our establishment in 1968 published full thermal performance characteristics for all our wheels for balanced counterflow. These can be down loaded from our web site as characteristics performance tables for all 3 series of rotor matrix porosities that we manufacture. We provide this as a useful tool to help building air conditioning design practitioners develop more sustainable designs.


We have recently introduced a new rotor variable porosity series into our range of heat wheels. This new variable series continue to have the same design and overall dimensions of all our other series S426, S430 and S436. These rotors are custom built to a selected porosity between the existing two extremes of S426 (low porosity, more Mylar, higher efficiency at higher pressure drop) to S436 (high porosity, less Mylar, lower efficiency lower pressure drop).

This is achieved by varying the small spacing between the Mylar layers to better match the specific design performance of efficiency and pressure drop of a particular project. This effectively increases or decreases the Mylar surface area for heat transfer per wheel diameter.
Bill Ellul
Managing Director B.E (Hons), FIEAust, MAIRAH



Download File


Download File
Revised 30 Oct 2014 Copyright © 2014 Bill Ellul. All rights reserved