Discussion

I have a query and your feedback would be greatly appreciated - we are looking into the reasons why two cast iron boilers packed up after 8 years of installation. The evidence suggests that there have been condensation at the back of the boilers. The system employs a primary pump set, a header (supposed to act as a low loss header), and secondary heating zones with each with its circulation pump and the mixing valves for weather comp. I am aware that CIBSE guide H mentions about figure 5.23 which eliminates the need for a back end protection for high efficiency modern boilers which is similar to what has been installed. However, the secondary system is a one-pipe system and there have been many leaks over the years - the header is not a vertical one but horizontal and it is 125mm for 37.5 m3/hr flow rate and I believe this is undersized which is in effect acting as a piece of pipework and bringing the cold water back from the system when the system is heating up. I am looking to overcome this issue with the condensation by installing a single pipe after the main pumps before the LLH with a two port valve and a thermostat set to say 60'C when lower than that the two port will open and mix the primary with the primary return to the boiler. I can also consider a bypass with a three port mixing valve if a better solution but this would be a more expensive option. Any suggestions would be greatly appreciated. Thanks.

Hope this is of some use.

I would also say a 3port valve is better as it guarantees that you can completely stop if necessary and load water coming back to the boiler.

I would also ask you does the cost of a 3port over a 2 port really justify putting another pair of boilers at risk of early failure?

Installing a 3-port shunt valve is a classic solution to the problem, however, one could also undersize the primary circulator to buy a longer run time of the boiler to compensate for the return water temperature.Good boiler manufacturers include a dedicated low temperature return circuit as a percentage of load to avoid such a problem.

You didn't say why the boilers failed, and what the mode of failure might have been. Boilers fail primarily, due to either stress, or corrosion, and the determination of the failure will inform the necessary solution.

In earlier times, when sectional boilers were installed burning coal, or worse, high sulphur heavy fuel oil, then fire-side corrosion could be a problem, especially with heavy fuel oil installations. Such plants were ( usually ) shut down overnight, and suffered from long periods of operation under condensation conditions, during the daily start-up. Sulphur dioxide and trioxide, produced during the combustion process combined with moisture to form sulphurous and sulphuric acids, which caused severe corrosion of the boiler metal. The problem first came to notice with the conversion of coal fired plant to oil firing, in the 1960's and 70's, with the hitherto unknown early failure of boiler sections that had previously been quite satisfactory on solid fuel installations. Coal fired plant normally operates continuously through the heating season, and thus was not generally subjected to the repetitive cold starting of oil-fired plant. Also, coal has a much lower sulphur content, than some of the residual oils, which were common-place at that time. The phenomenon was characterised by the final sections of the flue passes wasting away, and was given the term "back-end corrosion".

The more enlightened of us realised the problem, and it was noted that there was a dip in the rate of corrosion, if the boiler heating surfaces could be maintained above 60 degC.

Thus began the use of Shunt pumps, and a variety of other suggested solutions, including internal mixing arrangements, all designed to raise the temperature of the water in the boiler to this magical temperature, in the shortest possible time . . .

There was much debate as to what size the shunt pump ought to be, and how it should be controlled. Frankly, most of it was shite, and it says much about the members of our auguste organisation, that so many old-wives tales persisted !

Simple thermodynamic analysis provides the necessary guidance on the matter . . . .

From a cold start, the boiler has to heat up several thousand litres of cold water in the heating system, quite apart from dealing with the intended emissions from heating surfaces of radiators, etc. It's quite easy to work out an approximation of how long this might take, and how long the boiler might be subjected to corrosive conditions.

The use of a shunt pump, in any shape or size, does nothing to improve the situation, and thus was, and is, a waste of time ! In those days, shunt pumps had a different, and useful purpose, in that they maintained a minimum flow-rate through the boiler under low-load conditions, usually when the heating mixing valve was closed to the hot port, and stopped the problem of kettling and nuisance tripping of the high temperature themostat.

The proper way to provide back-end protection, as it was known, is by a three-port valve, arranged to divert the boiler flow water directly back into the boiler return pipework, and a separate primary pump to maintain flowrate through the boiler during its heating up period. The three port valve should have a sensor in the return pipework, downstream of the common ( mixed ) port of the three port valve. This sensor should be set to give a controlled temperature of 60 degC, and an integral, or floating controller will provide the necessary control action. The heating system will not heat up any more quickly, but the boiler will reach a 'safe' working condition, above corrosive dewpoint, in a very short time. This is the correct and only satisfactory way of preventing corrosive failure from combustion product condensation.

But is that what you are witnessing ?

Gas fired boilers are similarly subject to condensation during start up, and the dewpoint of the combustion products at stoichiometric air/fuel ratio is 55 degC. Thus the need to achieve a minimum heating surface temperature above this, in order to avoid condensation. Except, of course in condensing boilers, where our objective is to be below condensing temperature as much as possible . . .

Although the condensation is acid, because natural gas contains negligible sulphur, it is only mildly so, and the condensation does not generally pose a serious problem to the boiler metal. It is rather more a nuisance, leaking out of the boiler and over the boiler house floor.

So, is the failure likely to be due to corrosion, or might there be some other factor, and the condensation a 'red-herring' ?

Well, as previously suggested, boiler sections (and steel boilers ) also fail from stress. Generally this is due to overheating of the heat transfer surfaces, or unequal expansion of various parts of the boiler. Dealing with the latter first, on sectional boilers freedom of movement is important to allow expansion to take place. The use of tie-rods, running the length of the boiler, to hold the sections together usually incorporates some form of expansion provision to allow movement. I have seen fixing lugs torn off the front and back sections, where such provisions have been omitted or assembled incorrectly. Similarly, rigid short pipe connections may constrain the boiler and prevent its free movement. Such failures are fairly obvious, as cracks etc. on the outside of the boiler sections, and the solutions are also pretty straight-forward.

Now thermal stress is much more difficult to pin down, and more so, the cause. Failure here is more likely to be within the boiler, and requires deeper analysis.

The design of the boiler, and the way in which the gases travel through the convection passes will have a bearing upon the likely cause. The thermal loading of the combustion chamber will help to indicate what might be going wrong. Divide the burner thermal input by the volume of the combustion chamber : a figure of less than 1200 kW/m3 for reverse-flame boilers and 1800 kW/m3 for conventional three-pass boilers is a good starting point. A length to diameter ratio of 2.5 : 1, for reverse flame boilers and 3.5 : 1, for three-pass boilers will also generally be acceptable.

Try to avoid very short or maximum length sectional biolers with a reverse flame configuration.

Is the boiler being fired at the correct rate ? and if it is range-rated, where in the range is it ? The conclusions should be obvious.

Is the burner compatible with the boiler furnace configuration ? The type of flame which is produced suits a particular arrangement, and we need to tailor our flame pattern to suit the specific boiler. Some burners simply will not fire a particular boiler furnace successfully. Burners on a reverse-flame boiler have the most challenging task : The burner must have a high enough velocity across the diffuser head to ensure that the flame travels all the way down the furnace, before turning back on itself, and so avoid the possibility of short-circuiting, but not impinge on the back wall of the combustion chamber, with the attendant possibility of overheating the back section. Early turn-around and short circuiting of combustion products will mean a high temperature on the front boiler section, and often the early failure of the front door refractory. Combustion gases entering the convection sections should not exceed 950 degC, if you are able to measure it.

So, what else might there be in the way of causes ? Well, we could consider the water side. Have you checked that the design water flowrate is being maintained through the boiler, under all conditions ? Have you done a water analysis, and got some meaningful results ? Does it predict the formation of scale on the waterside of the heat transfer surfaces ? Check out the presence of silica based scales which, in some cases, require only fractions of a millimetre thickness to seriously impede heat transfer and cause overheating. I once witnessed two 3.5 MW reverse flame boilers fail in just over twelve months from installation, due, in part, to the presence of such scale.

Has that given you food for thought ?

Good luck !

Steve A.

We have another project where a colleague chose 3 steel boilers (ideal Vanguard range). There is a pair of primary shunt pumps, low loss header arrangement serving a large school with 4 zones. There is quite a lot of condensation happening at the back of the boilers and I am asked to provide a solution to prevent condensation happening (why me!! :). Before doing anything such as back end protection which the burner manufacturer was suggesting I was thinking in the lines of if I could bring on the boilers and the primary pumps first to bring the low loss header and the primary to the design temperature first thing in the morning before the secondary pumps come on this would give the system a chance to reduce the time that the boilers are exposed to the condensing temperatures during the start up period. If this doesn't resolve the issue I was thinking I can start the secondary pumps at intervals to keep the return to the boilers above 55'C. I was thinking - hopefully not complicating the matters too much- I will have to work out total volume of each zone and knowing what we can deliver i.e. flow rates and kW rating - I may be able to work out the time it will take for the zone to get up to the temperature etc. However as it will end up in the Low loss header it will mix up with the boiler flow and not sure how to work out the temperature coming back to the boiler return. Did anyone do anything similar to this? I know there is a risk of not being able to hit the design conditions early enough in the morning so I will make sure this is accounted for. thanks alot. sorry I took so long!