Heat exchanger basics
This particular heat exchanger (HE) is the latest incarnation from Giannoni and is utilised on many fully premixed condensing boilers. The donor HE (from a Baxi Platinum) was replaced by the manufacturer under warranty, as it had suffered an internal leak after approximately one year.
The heat exchanger comprises an outer thermoplastic casing. Within the outer casing 6 bundles (comprising 4 coils each) of stainless steel oval formed tube are coiled. Lower thermal output heat exchangers may have fewer bundles.
The rear most bundle is axially separated by an insulating refractory fibre disc and stainless steel dish, from the front set of tube bundles.
A stainless steel heat shield surrounding the coils serves to protect the plastic outer casing from excessive heat.
Two headers are mould ed into the outer casing. These serve to connect the coil ends in the correct pattern to allow correct flow of water through the tube bundles. Two header caps are then utilised to form flow and return connections (on one side) and a common rail connection on the opposing side. The flow and return header cap utilises a moulded web forming a division in the centre. This allows separation of the flow and return water paths into their respective coil bundles.
A thermistor pocket is formed at the back of the front section of the outer casing. The thermistor projects through the heat shield and senses the flue gas temperature at the periphery of the coil bundles. It forms part of the boiler safety system and shuts down the burner if excessive temperatures are reached due to dry firing/significant heat exchanger fouling/poor thermal transfer etc.
A condensate outlet spigot is moulded into the lower section of the outer casing at the rear. Flue gas is ported through the rear of the casing into a vertically formed duct to which the flue is subsequently joined.
Operation
The basic principles of operation are as follows:
A tubular stainless steel burner is positioned centrally within the front section of the heat exchanger. A front cover plate and gasket assembly forms a gas tight seal at the front face. This cover is secured to the heat exchanger by retaining nuts fitted to the end of the stay bolts (see later). High pressure combustion products emitted by the burner are forced radially through the front section tube bundles. The flue gas (having now lost significant heat energy) passes along the periphery of the outer coil face towards the rear of the heat exchanger.
Gas flow is prevented from passing directly to the flue gas outlet by the insulating disc and stainless steel dish. The flue gas is forced back through the final rear section tube bundle towards the centre of the rear section. The combustion products now pass through the rear circular exit port and then vertically towards the flue outlet spigot. In this process water vapour is condensed on the rear tube section and drains via the condensate spigot. Water vapour condensing in the flue has a pathway back through the heat exchanger outlet to the condensate drain. Small holes in the lower heat shield half, allow condensate to drain though to the condensate outlet spigot.
Details
With reference to each image a more detailed description of the heat exchanger construction now follows:
Image 1 (see here)
This shows the complete heat exchanger. The outer plastic casing is formed in two halves. A finger or comb joint is formed between the two halves This joint is bonded with a high temperature silicone sealant and additional small screws. The front stainless steel end plate is crimped around the circumference of the plastic casing halves. A silicone based sealant forms a seal. The end threads of each stay bolt can be seen (retaining nuts that pull the stay bolts tight against the front end plate have at this stage been removed). Inside the coil bundles can be seen. At the rear the insulating disc of refractory fibre has been removed (it is normally retained by a penny washer and nut). The insulating disc prevents heat transfer to the rear section. The lower header can be seen. The front spigot is the flow connection ; the rear being the return connection. These joints are sealed by conventional O ring and retainer clip. The condensate spigot can be seen at the rear.
Image 2 (see here)
This is a more detailed look at the lower header section. The main heat exchanger shows the six tube bundle ends terminating at the header. Looking carefully at the header cover (comprising the flow and return connections) you will note the formation of a central division web. This allows return water to flow through the rear 3 tube bundles to the top header and then back through the front 3 bundles to the flow connection. A single O ring comprises the main seal of header cap to casing. A silicone sealant is then applied to form a secondary seal. The cover is retained by 12 Torx screws.
Image 3 (see here)
Here we see a plan view of construction. The interlocking finger joint can be seen at the lower edge. Note this joint is not as manufactured. Disassembly required the joint to be cut. Just visible is the lower heat shield half. The stainless dish separating front and rear sections is also shown.
Image 4 (see here)
We are now viewing the heat exchanger from the rear. The rear face of the separating dish can be seen. On the right one of the two stay bolt arrangements can be seen. At the rear each two stay bolts are welded to a rectangular section stainless bar. Raised nibs are formed around the circumference on the flattened section of each coil. In this way as the stay bolts are tightened the coil bundle is compressed. The raised nibs allow a controlled gap between each coil. Reducing this gap to a minimal setting increases the thermal transfer of combustion products to the coil surface by minimising boundary layer effects.
Image 5 (see here)
We are viewing the rear top half section of the heat exchanger from the front. This is actually formed from several mouldings some screwed together. The overheat thermistor pocket is shown on the right. Each bundle connection to the top header can be seen. To enable disassembly each connection had to be cut.
Image 6 (see here)
A plan view of the top heat shield half. The cut section of each bundle is shown. Note heavy discolouration and residues from such a short time in service. A flue gas sensor hole is formed in the heat shield allowing projection through to the coil bundles.
Image 7 (see here)
Here we see a detailed cut away section of the header arrangement. After the tube bundles have been inserted through the header sections the tube ends are hydraulically swaged to form a correct O ring joint. The groove for the main header cap O ring seal is also shown.
Image 8 (see here)
This image shows the front tube bundle stretched out to show the impressive tube forming. The small raised nibs on the flat section of the tube can just be seen. The lower heat shield is allowed to drain condensate via holes. The two left hand stay bolts can just be seen.
Build up
The design raises a number of questions (especially with long term reliability in mind) and in no particular order:
Very complex tube forming techniques are utilised especially as the oval section transforms to the circular section. Will stresses induced during this process lead to stress corrosion cracking in later life? Each coil is approximately 2.5 meters in length. Will the cycled thermal expansion lead to stress cracking in both metal and plastic parts? The plastic outer casing and metal coils have considerable differing thermal expansion coefficients.
The raised nibs are continually in contact with their opposing coil surface. Staining was evident around these nib contact points. Will these closed areas promote crevice corrosion? The steel is likely to have a reduced thickness as the material is stretched to form the nib.
The stainless steel oval section measures approximately 25 x 7mm with a wall thickness of 0.7mm. This is a significantly thinner wall thickness than say the Viessmann heat exchanger coils. I would contemplate the resistance to both crevice and pitting corrosion is substantially reduced.
Cleaning the heat exchanger will only have minimal effect since full access to the coils (in particular the rear section) is not possible. The lower heat shield also allows debris to accumulate.
O rings are utilised for both the header and each tube end to header connection seal. O rings are a much abused device within the heating sector. System water chemical composition can have a drastic effect on seal life (especially if contaminated with mineral oils). Oils used in the coil forming and swaging operation may contaminate the O rings. Plasticisers leaching from the O ring to plastic components (and vice versa) or mould release agents can cause deterioration. Only minor overheating may seriously impair O ring life. As we know many boilers apparently well installed suffer from O ring embrittlement after only a few years in service.
A further common problem is poor moulding tool maintenance. It is not uncommon to find a slight flashing line found where the mould tools part. Although appearing insignificant it often leads to slight water leaks. A simple quality problem is often overlooked during manufacturing. Since access to remove the flashing is impossible after manufacture this “minor” manufacturing fault would now render the unit unserviceable.
Insufficient system cleaning and lack of chemical inhibitors will allow debris accumulation within the coils. Conventional copper heat exchangers are formed from horizontal sections of tube. Debris build up tends to be a slow process. The coiled tube design set in a vertical plane of this heat exchanger allows debris to settle in the base of each coil (especially after boiler shutdown). Three bundles are in parallel with one another. Therefore debris could accumulate in an individual tube whilst apparent satisfactory flow continues through the others.
Eventually a bundle may completely block. It is probable this fault would show up as a flue gas overheat shutdown since heat energy is not sufficiently extracted from the flue gas. . Cleaning may be difficult since flow would always be through the easiest pathway. Single pass tube designs such as the Viessmann and Buderus heat exchanger will have a higher fluid velocity and should be far less prone to this problem (although internal wall scouring may then become a problem!).
Control
The heat exchanger shell is manufactured from a variety of plastics. The headers are moulded from a blend of Polyphenylene and High Impact Polystyrene. For reinforcement 30% glass fibre is added. This mix tends to be very hydrophobic and so retains high dimensional stability. The outer casing is a slightly different composition and appears to be a blend of Polyphenylene Ether, Polystyrene and Polypropylene. Again further reinforcement is provided by a 30% glass fibre inclusion. This blend appears to offer a little higher temperature performance.
Again long term reliability is difficult to judge and although plastics have evolved considerably over the last few decades many of us have witnessed serious plastic embrittlement during boiler repairs - perhaps due to system chemicals and prolonged heat. Although these materials are available with a fire retardant property it may surprise you to find they are actually very flammable once ignition temperature is reached. In fact both materials burn at an alarming rate producing extremely sooty (and noxious) combustion products much a kin to burning acetylene in the absence of oxygen.
In view of this potential hazard the flue gas temperature sensor is of utmost importance. However, the thermal response from a thermistor bead type sensor is fairly slow and it may prove marginal in the least in preventing serious overheating of the plastic components during major boiler malfunction. Primary safety detection systems such as static and dynamic water pressure sensors will help to combat this fault condition but whether the test houses require this additional protection I have not investigated.
On the Baxi Platinum for instance only a minimal protective system is employed - a simple on/off pressure switch (CIME) is incorporated. This however, will not detect lack of water within the heat exchanger ie air, reduced /zero water flow from a partial/full pump failure or blockage. Note, any flow pipe temperature sensors and overheat thermostat sensors will produce no useful indication of excessive temperature rise within the coil bundle if there a lack of water flow since the heat exchanger pipe-work is effectively insulated from the flow pipe-work via the plastic header. Where return thermostats are incorporated the rate of change of temperature can be measured to provide useful additional protection and correct delta T‘s matched to burner input.
A linear output pressure sensor (ie to measure the pressure increase/decrease on water flow) or a mechanical diaphragm responding to differential pressure would also provide a higher safeguard since water flow will be detected. Some pressure sensors notably the SIT model is very prone to blockage (often due to poor sighting of its inlet port) and the plated contacts have poor corrosion resistance (they are exposed to the atmosphere through the balancing port at the rear). It is conceivable the pressure switch could indicate sufficient pressure even on a drained system.
Although manufacturers must enable their products to be re-cycled this will not be particularly straightforward on this design. Since the casings are bonded together with high strength sealant and the bundle pipe ends are swaged disassembly is not easy. It took approximately 1.5 hours using hand tools and much hard work. No doubt mechanised tooling would significantly reduce this labour, however the awkward design and flexibility of the individual components makes disassembly hazardous. Consequently landfill is a very likely prospect at end of life. On the contrary, copper based heat exchangers (generally very easy to remove) are highly prized items and are routinely re-cycled.
And also due to the manufacturing techniques employed, repairs for example, to a simple leaking O ring are not possible ie a minor fault renders the unit scrap.
Conclusion
In conclusion the design of the heat exchanger allows a very compact size versus high heat output performance. However long-term reliability must be questioned since on initial inspection many design features appear a little marginal. The nature of the industry shows customers reluctance to part money for quality installations (higher expense in part from legislation) and an increasing number of poor quality installations (especially in our inner cities with a low cost labour force); this may lead to a very short heat exchanger life. The majority of warranties currently offered by manufactures exclude all claims resulting from system debris/poor installation. As usual only time will tell.
