A respected colleague of mine once stated a rather solid control principle which I still support to this day, however conditionally so when it comes to combustion control with steam boilers. What he said was this: the most effective control is achieved with a direct (as opposed to an indirect) approach.  Now let us look at and explain this principle in light of boiler control, specifically with coal burning boilers.

Of specific concern to users of steam boilers is the steam pressure, as it is an indicator of the difference between the steam demand and the current rate of steam generation. A dropping steam pressure, for example, indicates that steam demand exceeds steam production, and this is the signal used by the control system to regulate the combustion response to meet the increased steam demand. It does so by feeding more fuel to the furnace (stoker speed increases), which in turn requires more combustion air (FD fan speed increases), and also increase of the ID fan speed to ensure the furnace pressure stays below atmospheric pressure at all times. Yes, positively no smoke allowed inside the boiler house! It is toxic stuff.

Thus the control system has to perform a delicate balancing act: feeding more coal to produce more steam; supplying more combustion air to meet the oxygen demand of increased fuel feed; and increasing suction by the ID fan to keep the furnace pressure in check.

The big question for many years now boils down to this: upon a change in steam pressure (steam demand), which one of the combustion components should be the first to respond? If we assume a drop in steam pressure, then according to my colleague’s principle of direct control, it means that the stoker should be the first device to respond (known as stoker leading control), seeing that more steam is produced by feeding more fuel. The control sequence in this instance is as follows: when the steam pressure drops, the stoker speeds up (to fuel the fire to produce more steam); the FD fan speeds up (to increase combustion air to the fire, but this also pressurizes the furnace); the ID fan speeds up (to balance the increased furnace pressure).

But there are a number of negatives associated with the “direct control” approach in this instance:

  1. The ID fan, because of its greater inertia, may be slow to respond to changes in furnace pressure and furnace pressure may turn positive, even just momentarily, and cause some smoke discharge into the boiler house. To cater for the sluggishness of the ID fan the rate of response of the control system to changes in steam pressure may have to be set rather slow so as to avoid the furnace pressure from fluctuating too widely around the set point
  2. Controlling the furnace pressure by means of the ID fan wastes some electricity because of the inertia of the heavy fan impellor, which is continually adjusting its speed to maintain furnace pressure.
  3. The big fan speeding up and down all the time in its quest to control the furnace pressure is clearly audible in most instances and can become an annoying aspect of boiler operation.
  4. The biggest challenge with stoker leading control however is when it is used with a twin flue boiler equipped with a single ID fan, two stokers and two FD fans. With this configuration and for stoker leading control to function properly both stokers are required to operate at the same speed and to respond to changes in steam pressure in the same way. Consequently both FD fans have to operate at the same speed too to provide equal quantities of combustion air to the fire on each stoker (assuming FD fan curves are identical). The ID fan controls the furnace pressure from any one of the two furnaces, or at least from the one with the highest pressure. In theory thus the air-fuel ratio of combustion can be controlled fairly accurately in both furnaces.
  5. In the practical situation however we often encounter substantial differences in furnace pressure, even with FD fans of both flues operating at the same speeds and delivering equal quantities of combustion air. For some reason or the other the resistance to combustion air and flue gas flow per furnace is seldom identical under (perceived) identical combustion conditions. During a recent visit to a user of a steam boiler I found the left hand flue operating at -20 Pa furnace pressure, and the right hand flue at -56 Pa with FD fans and stokers operating at similar speeds, left and right. If it wasn’t for the high volatile coal being used they may have experienced problems with the right hand fire losing ignition because of cooling of the ignition arch. Ingress (tramp) air on the right hand side will also be higher, eroding combustion efficiency without the user knowing it. 

An indirect control approach eliminates some of the drawbacks of the stoker leading concept. It is known as an ID fan leading approach. In this instance the ID fan is the first device to respond to changes in steam pressure. Although it does not feed fuel directly into the furnace, it initiates a sequence of events leading up to an increased fuel feed rate. The control sequence in this instance is as follows: steam pressure drops; ID fan speeds up (and furnace pressure drops); FD fans speed up (to balance furnace pressure and to increase combustion air flow); stokers speed up (to fuel the fire for increased steam production). The positives associated with this “indirect control” approach are as follows:

  1. Furnace pressure is controlled by the FD fan. Responses are generally faster, less audible and more energy efficient.
  2. With twin flue boilers furnace pressures are individually controlled. Although this implies that each set of FD fans and stokers will operate at a different speeds, it still offers the opportunity to control each furnace as an individual single flue boiler, except that they share a common ID fan, the speed of which is primarily determined by the steam demand.

Let us however not be fooled into thinking that combustion control is a simple matter, and so much the more with a twin flue boiler which fires pea coal on chain grate stokers, and a control system which is efficiency based. Although I personally favour the ID fan leading type of control, it is not a perfect control approach and presents its fair share of challenges which may be addressed in future Boiler Bits.

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.


There is a portion of scripture referring to little foxes ruining the vineyard. Along the same vein I have spotted a number of little foxes ruining boiler operations as well, often to the utter confusion of operating staff and the dismay of Production staff. One of them in particular is simple to avoid, but seemingly difficult to detect. I am referring to the draught sensing pipe running from the ignition arch through to the control panel. It is normally a ⅜ inch (10mm) copper pipe which is run along a cable tray, with a number of pipe joints and connections along the way. Quite often the copper pipe is joined to a plastic or rubber pipe on its final run to the control panel, where it is connected to a furnace pressure controller, such as a photohelic or other electronic pressure differential sensing device.

One thing is for sure: if draught (furnace pressure) sensing is lost, boiler operations tend to end up in chaos. Either the boiler house rapidly fills up with smoke and flames, or the FD fan and stoker come to a standstill. Both are of cause equally disruptive; on the one hand choking operators trying to engage manual operation to restore steam production, on the other hand dropping steam pressure and irritated production supervisors swarming and/or calling the boiler house to find out why there is no steam.

I walked into our boiler house the other day after a 36 monthly statutory inspection. I noted that the draught sensing pipe has been replaced, but I also noted that the pipe sloped slightly upward where it took the first turn. Not good at all! With a freshly cast ignition arch I knew there was a significant amount of water evaporating from the wet refractory and condensing in the sensing pipe. Very soon the condensate would fill the pipe at the first bend and block the furnace pressure signal to the control panel. My suspicions were confirmed when I disconnected the pipe at the first joint and pressed down on the loose end – some 5 ml of condensate drained out on to the floor.

So here are a few practical tips to consider regarding draught sensing pipes and eliminating the little furnace pressure fox:

  1. Make sure the sensing pipe slopes downward all the way to the panel, but install a deliberate loop where condensate may gather and which is easy to disconnect and drain. I have even seen automatic condensate drains to be installed on these loops.
  2. Condensate only really becomes a problem when a new ignition arch is cast. Regularly check for the presence of condensate in the sensing pipe during the first two weeks after the arch was cast.
  3. Reduce the number of joints and connections to the minimum. Each one of them is a potential source of leakage which can disrupt the quality of the furnace pressure control signal. Ensure joints and connections are tight, well supported and generally in good condition.
  4. Avoid long lengths of sensing pipe. We once had to do a sensing pipe run of 60 metres long, where the control panel was located in a remote room. We could spot a definite delay in the response of the system to changes in furnace pressure. Eventually we could improve the situation by increasing the diameter of the sensing pipe, but it was everything but ideal. It is probably better to run long lengths of control wiring than sensing pipes.
  5. Be careful when using plastic sensing pipes which may be exposed to high temperatures. We once encountered a situation where someone welded on a steel structure in the boiler house and a droplet of weld spatter burned a small hole through the plastic pipe somewhere on a cable tray. It took us almost a day to find the cause of the erratic draught signal.
  6. Furnace pressure control going wrong can cause significant damage, especially if it turns positive in the absence of a boiler operator. Consider installing protection that will trip the boiler when furnace pressure turns positive for too long – a feature that may one day spare one the cost of expensive repairs.

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.


We recently did work for a Client who occasionally fires sunflower hulls with his fire tube boiler; at other times he uses pea coal. We used the opportunity to provide a simple change-over capability with his boiler control system so that either coal or hulls firing can be selected with the push of a button on the operator panel.

At the time of commissioning the control system alterations I noticed that the operators operated the boiler at deep (negative) furnace pressure (-45 Pa) and extremely high stoker speeds to prevent the fire from burning back into the fuel hopper. To me it seemed that the deep negative pressure in fact promoted burn back, rather than preventing it, especially at lower stoker speeds.

So our first step was to increase furnace pressure to -10 Pa and, lo and behold, the burn back problem improved dramatically. However, a new phenomenon surfaced – with the FD fan and stoker running at the same speeds as before (ID fan now running slower), the length of the fire increased significantly, with some burning fuel dropping with the ash into the ash trolley. The only reasonable explanation for this new problem was that less (tramp) air entered the furnace now that the furnace pressure (draught) was more positive than before, resulting in less available oxygen and slower combustion of the fuel.

Going back some 10 years into the past I remember an incident where a client complained about his boiler smoking excessively under normal operating conditions. I happened to be on site at the time and suggested he lowers the furnace pressure from -5 Pa to -20 Pa. The shade of the smoke from the stack turned significantly lighter, much to the delight of the user who was desperate to do something about his smoking stack. I later realized that most of the tramp air introduced in this way probably acted as over-fire air, mixing with the volatiles, increasing turbulence and providing for more complete combustion of the combustible gases.

With this in mind I conducted an experiment on a brand new coal fire tube boiler for which we supplied the control system. The client was under immense pressure from the local community to reduce smoke emissions from the stack (coal volatiles exceeded 30%) and hence the boiler was operated at elevated levels of excess air and deep draught as a means to alleviate excessive smoke and protecting the guillotine door from fire burning back under it.

With a portable gas analyzer in hand I conducted a simple experiment to establish the influence of draught on flue gas oxygen level and hence combustion efficiency. With the boiler operating under stable conditions the draught was set at -5 Pa and the flue gas O2 level was measured at 10,6%. Thereafter only the speed of the ID fan was increased to lower the draught to -25 Pa and the flue gas O2 level increased to 11,5%. No adjustment of excess air, or anything else that could influence the air-fuel ratio. But even so the combustion air increased by almost 10%! And even with a brand new boiler tramp air ingress seems to be a draught induced reality.

So what is the significance of all of this? Probably more than one may realize at first glance. Firstly my combustion calculator indicated an additional 2% fuel energy loss because of the lower furnace pressure and resulting increased flue gas oxygen percentage.

Secondly one can only speculate at the response of the control system if a flue gas analyzer has been incorporated to trim the air-fuel ratio – most tramp air bypasses the fire, or fails to take part in the combustion process. Read more about that in Boiler Bits 8.

It is not unusual to see boilers being operated at -20 to -30 Pa furnace pressure in this day and age, especially those boilers still equipped with photohelic type draught controllers.

Want to save fuel? Consider the furnace pressure at which the boiler is being operated. Quite often some of the low hanging fruit is not so easy to spot, but easy to find if one knows where to look. And don’t underestimate the benefits of investing in a high tech combustion control system – it may reveal even more low hanging fruit!

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.


Continuous water level control is gaining popularity among users of steam boilers. This particular technology uses a water level transmitter (usually a capacitive probe) installed on the boiler to measure the water level. The level signal from the transmitter is used to control the speed of the feed water pump, which is driven by a frequency inverter. The speed of the feed pump is thus controlled by the water level – the lower the level drops, the faster the pump will run and the higher the rate at which water is delivered to the boiler will be. And as the water level rises, the pump will slow down until it may eventually stop.

There are several benefits of this technology over the established practice of stopping and starting the feed pump from a level switch:

  1. Reduced wear of mechanical components and electrical switch gear due to pumps starting and stopping.
  2. Elimination of steam pressure drop and thermal stresses while the feed pump is filling the boiler with relatively cold water at maximum delivery rate, repeating the process many times over per day.
  3. Water is replenished to the boiler at the rate at which it is being evaporated. No detectable pressure drop or cyclical thermal stressors.
  4. The continuous water level control system adds a back-up safety device for the protection of the boiler. The signal from the level transmitter can be used to sound a siren and to trip the boiler when the boiler water level approaches dangerously low levels. It can also warn against excessively high water levels and can even stop the feed pump under these conditions.
  5. In terms of power consumption it is not easy to come to a conclusion as to the best pumping configuration – there are just too many variables determining how the feed pump is required to operate. However, my gut feel is that the continuous water level control configuration may consume more power than its on-off counterpart. The reason for this is that the variable speed pump ideally has to operate continuously and maintain a relatively high speed, even at minimum delivery, to overcome the resistance of the steam pressure acting on the boiler water.

Setting up the water level control algorithm is however rather tricky and may even be baffling to the uninitiated. Let me explain: under normal operating conditions the water in the boiler is boiling off at a certain rate to satisfy the steam demand. The water mass thus consists of a mixture of saturated steam (vapour) bubbles rising to the water surface, and liquid water. This mixture of water and steam bubbles fills the boiler to a safe water level.

As soon as “cold” water is being pumped into the boiler, it cools down the existing water and vapour mixture, causing the vapour bubbles to collapse (condense) and the water level to drop rapidly – a phenomenon known as shrink. So although the pump may be pumping water at maximum capacity, the water level may be seen dropping for a while, after which it stabilizes and then starts rising. As soon as the set point water level is reached, the pump has slowed down to minimum speed, but at that point the “cold” water has been heated to boiling point and it starts boiling more vigorously, vapour bubbles increasingly develops and the water level may be seen to rise steadily – a phenomenon known as swell, even though the pump may not be delivering any water.

Note that a rapid drop in steam pressure will also trigger the swell phenomenon, as a reduction in pressure will increase the rate of evaporation of the saturated water mass. On the other hand, if the steam output from the boiler is reduced suddenly, the pressure will increase and cause the water level to shrink as the rate of evaporation decreases.

Rather challenging if one tries to control the water level without knowing what is happening on the inside of the vessel.

I am not going to involve myself with the various methods of water level control here – solutions range from elementary to extremely complex, especially with water tube boilers with relatively small steam drums. Maybe this may form a topic for a future publication.

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.


You are probably familiar with the situation – a user of steam boilers returns a low percentage of condensate to the hot well, and is concerned that the temperature of the feed water is relatively low. He somehow knows that boiler efficiency is linked to the temperature of the feed water, and it just seems so logical that heating the feed water would increase efficiency – a fairly low cost high yield intervention.

A few elementary calculations, based on a number of realistic assumptions, reveal that a 40 ᵒC feed water temperature increase may increase boiler efficiency by almost 12%, or in terms of fuel spend the bill will be 12% lower than before. Sounds too good to be true? Well, it probably is.

Engineers battling every day to cut steam cost to the minimum know how much it takes in terms of management, supervision, operation and technological input to improve efficiency by a single percentage point. Twelve percent improvement is huge in any engineer’s language.

On two occasions during the past six months I engaged in debate with engineers regarding the perceived benefits of steam heating of the hot well, and to this day I am not sure if my counterparts have been convinced by what I have said. Maybe the time has come to put pen to paper and to analyze the situation in the light of simple logic.

Let us assume a scenario where a boiler is operated at 10 bar (g) pressure, using feed water at 40 ᵒC from the hot well. The steam demand is 1000 kg/h under these conditions. The engineer decides to inject steam into the hot well to raise the water temperature to 85 ᵒC, aiming to improve boiler efficiency and to save fuel in doing so. A steam heating system is installed to the hot well and put into operation.

For the hot well to be heated to the target temperature of 85 ᵒC steam energy to the tune of (4.2 kJ/kgᵒC*40 ᵒC*1000 kg/h) = 168000 kJ/h is required. Assuming an overall thermal efficiency of 70% the quantity of steam required to meet the heating requirement = (168000 kJ/h/2000 kJ/kg/0.7) = 120 kg/h. Although the feed water temperature has now been raised to a level where 12% efficiency is gained, the steam consumption has also increased by 12% to 1120 kg/h to provide the additional steam required to increase the feed water temperature. So whatever fuel savings were gained by the higher feed water temperature were lost to the higher steam flow required to improve the efficiency.

We are obviously running in a circle here, which is to be expected, since the system is merely circulating energy within itself – energy is taken from the system at one point and introduced back into it at another.

Energy efficiency improvement can only be achieved if energy losses from the system can be prevented or reduced (e.g. steam leaks, stack heat loss, shell heat loss, etc.), or if lost energy from the system can be recovered (e.g. feed water heating in an economizer, condensate return, etc.). Sorry, no short cut to real energy savings!

But steam heating of the hot well is not all in vain and must be encouraged as a means to reduce oxygen in the feed water. As the temperature of the water increases, so the solubility of oxygen therein reduces and so the concentration of oxygen scavenger required to chemically treat the boiler water is also reduced. So by circulating energy in the system some water treatment cost savings can be gained.

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.


I am not often caught off guard in matters of boiler efficiency, although I must admit there are still phenomenons in this respect I find difficult to explain. But at least it creates a sense of expectation to know there is still more to be explored and discovered.

A few years ago we were called out to service a boiler control panel for a client. The panel was equipped with frequency inverters for the fans and stokers, a plc and state of the art oxygen trim control module, complete with flue gas oxygen analyzer. But a short circuit in the panel (caused by the client’s artisan) blew the output cards of the majority of the VSD’s and we were requested to reset the plc and control parameters. Within a few hours the boiler was back in service and producing steam, and everything seemed to back to normal.

A week later I received a call from a somewhat irate plant manager. His energy monitoring records showed that since our repairs to the control panel his boiler consumed approximately 20% more electrical power than before. Now this was a completely new one to me. If he complained about higher fuel consumption I could immediately think of at least five possible causes. But electrical consumption – this was completely unexpected. And I had no ready answer, as a matter of fact, I was dumb struck! I requested more information and he forwarded me the appropriate power consumption trends of his boiler during the past month. It showed a clear increase in electrical consumption since the time of the boiler being back on range.

I stretched my mind as best I could to recall what was actually done on the day of the service. And then it slowly started to dawn on me that we adjusted the oxygen set point. I remember that initially there was excessive smoke from the stack and we adjusted the flue gas oxygen set point upwards from 7,5% to 8,5% to clear some of the smoke. With the smoke problem out of the way we continued to complete the service. But could this have had a 20% impact on the boiler’s power consumption? – a mere 1% increase in flue gas oxygen level?

It was back to the drawing board for us, back to the first principles of fan laws and combustion calculations. So here we go, in a nutshell: 1% increase of the flue gas oxygen level means 22,4% increase of excess air and 8% increase of combustion air. According to fan laws fan power is a function of (fan rpm)³, which is (1,08)³ = 1,26. Mystery solved! Our 1% upward adjustment of the oxygen set point increased fan speeds by some 8% and fan power consumption by some 26%.

This incident really was an eye opener to me in a number of ways:

  1. We tend to focus so much on combustion optimization that we often forget that electrical consumption of boiler motors does form part of the total energy equation.
  2. Adjustment of any combustion parameter must be done with full consideration of its effect elsewhere in the system. I somehow get the impression that it is often just too easy to increase excess air to eliminate problems with smoke or stoker temperatures and nobody cares to find more novel solutions, or realizes how much electrical energy may be wasted in the process.
  3. Very few users of steam boilers measure or manage their boiler’s electricity consumption. The one client who did so, did me a huge favour in bringing the correlation between flue gas oxygen and fan power consumption to my attention.

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.


With coal still being the predominant boiler fuel in South Africa users of coal fired steam boilers are forever under pressure to improve boiler efficiency with the worst possible fuel source that demands constant supervision of the boiler and manipulation of combustion parameters to extract the best performance from the age old energy conversion technology.

Pea coal firing is not particularly kind to boiler efficiency. It possesses all kinds of nasty characteristics which work against achieving any spectacular efficiency. Variations in energy content, volatile matter, particle size, reactivity, ash content, fouling tendencies, and many more requires constant human intervention to extract as much useable energy from the system as possible. No wonder coal must be considered an integral part of the total combustion system when looking at ways to optimize the combustion process.

It is not unusual to encounter only superficial knowledge among engineers and operators regarding the characteristics of coal and its influence on the efficiency and capacity equation. To many coal is merely a black solid fuel and the cause of many combustion problems. But as long as it is A grade it is the best, they say. The irony is that, having analyzed many A grade coals, I have found the majority of supposed A grade coal delivered on sites is actually B grade, even as low as D grade! So let us put coal through the combustion process and see what is of general importance.

  1. The main purpose of coal is to provide the energy required for steam production. And if one of our primary objectives is to produce steam at the lowest cost it follows that the price we pay for (usable) energy is a telling factor in achieving efficiency goals. So it is always advisable to calculate the cost per usable GJ of energy purchased. And most often one will find that A grade coal renders the highest cost per unit of energy (although it may have other properties which make up for the cost handicap).
  2. Lower grade (CV) coal will always be higher in ash content and lower in carbon content. High ash content affects boiler performance negatively in that unburned coal loss tends to increase, while combustion temperature decreases and hence boiler efficiency is lower.
  3. Always remember that your boiler has been designed to produce a certain maximum steaming capacity (MCR = maximum continuous rating) based on a certain calorific value of coal. Burning C and D grade coals may inhibit boiler capacity, especially if a boiler is required to operate at or near MCR.
  4. The next bottleneck in the production of sufficient steam is the rate at which the coal ignites, which is determined by the volatile percentage of the coal. Fires running away are mostly caused by low volatiles. With low volatile coal the operator is forced to run the stoker at slower speeds and higher furnace pressure to allow sufficient time and temperature under the ignition arch for the coal to ignite. Aim to purchase coal with volatiles exceeding 23% to be safe.
  5. Unfortunately volatiles are also the primary source of smoke production in the furnace. Volatiles percentage approaching 30% is cause for concern in terms of smoke from the stack. First thing to check when your boiler suddenly starts smoking is the volatiles content of the coal.
  6. A very important aspect of combustion efficiency is the reactivity of the coal, which can be established by means of a petrographic analysis. The higher the reactivity of the coal, the faster it burns out and the less excess air is required for its complete combustion. On the other hand, the less reactive the coal is (higher inertenite content) the more difficult it becomes to completely burn out the coal over the length of the stoker grate. To speed up combustion and to avoid live coal with the ash the excess air percentage needs to be increased, and so does dry heat stack loss. Bottom ash losses also tend to be higher with less reactive coal.
  7. Check the size grading of the coal when it is delivered. The stoker grate has been designed for pea grade coal, which means coal pieces between 5 and 20 mm in size will burn well and burn out completely over its length. Too much duff/dust in the coal causes segregation and inhibits air flow through the coal bed, causing carbon loss in the ash, combustibles loss in smoke and even carbon loss through the stack. Large pieces of coal may be difficult to ignite and to burn out completely over the length of the stoker.
  8. Receiving a load of heat affected (altered) coal does occur occasionally. Without going into any technical depth, this coal will be slow to ignite and slow to combust, and in most instances it will be unsuitable for firing on a grate stoker.
  9. Then lastly one has to be aware of the fouling tendencies of the coal. This can be established from the ash content analysis of the coal. We know that phosphorus compounds are notorious for a phenomenon called bird’s nesting, which completely blocks boiler tubes. But there are other nasty elements as well, containing low melting point metals such as calcium, sodium, iron and others and which accelerates fire side deposition inside fire tubes.

Unfortunately managing coal procurement is no easy task. By the time the coal has been delivered to site and is being fired for the first time it is rather late for effective remedial action. Most users just push through with the batch of poor coal and hope for a better product with the next delivery. A few suggestions:

  1. Know your coal better than your supplier, or involve someone who knows.
  2. Find a trustworthy coal supplier with whom a good working relationship can be established – this may however take time.
  3. Coal suppliers usually present a proximate analysis of the coal they market. Check the date of the analysis presented. It may be of their best coal six months ago.
  4. The proximate analysis does not say anything about fouling tendencies, reactivity or ash fusion temperatures. Carry out more advanced analyses from time to time to verify these important characteristics.
  5. Insist on a proximate analysis with every coal delivery. Regularly verify supplier analysis against own or independent analysis reports.
  6. Consider acquiring and installing devices and technology to alleviate the effects of poor coal characteristics, such as swinging chutes, over fire air systems, etc.

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.


Although condensate return has nothing to do with combustion control, it does have a significant impact on the efficiency of the boiler. Over the past number of years I have been developing a combustion calculator and every now and again I am being confronted by the percentage of condensate being returned and its influence on the hot well (feed water) temperature. I am referring to a system where condensate is returned from the process to an open (vented) hot well at atmospheric pressure (0 bar (g)), and where it mixes with cold make-up water to replenish steam and condensate losses in the reticulation system and processing plant. The mixture of condensate and make-up is delivered to the boiler via a feed water pump.

Theoretically one can assume that condensate is at 100 ᵒC (assuming no heat loss from the reticulation system) and make-up is at ambient (±25 ᵒC) temperature. Let us further assume that 50% condensate is returned to the hot well and that make-up is 50% of steam supply. Simple arithmetic suggests that the temperature of the water in the hot well should be at 62,5 ᵒC.

With one of our own outsourced boilers I was however often confused by the obvious discrepancy between theory and the practical reality on this matter. Although the percentage condensate returned varied between 40% and 60% the feed water temperature was mostly in the vicinity of 94 ᵒC. Leaking steam traps in the plant? (The maintenance team denied it was the case.) Probably, but not necessarily so. Let us once again look at first principles governing the physics of steam and condensate.

With our outsourced boiler steam is supplied to the plant at 8 bar pressure and then reduced to 5 bar at the point of use. After passing through a process heat exchanger the condensate is returned to the boiler house. From steam tables we know that:

  1. The enthalpy of saturated steam at 8 bar is 2774 kJ/kg
  2. The enthalpy of saturated steam at 5 bar is 2757 kJ/kg. Since no heat is exchanged or lost in the PRV, the steam at 5 bar will be slightly superheated. This superheat is absorbed by the process in the heat exchanger.
  3. The enthalpy of saturated condensate at 5 bar is 671 kJ/kg. This is the energy remaining in the condensate after it has passed through the process heat exchanger, but before it has passed through the steam trap.
  4. The enthalpy of saturated condensate at atmospheric pressure is 419 kJ/kg. This is the energy contained in the liquid condensate after passing through the steam trap and on arrival at the hot well.

It seems that passing through the steam trap has caused the condensate to give up 252 kJ/kg of its internal energy? Where did it go to? Of course, it was used to create flash steam (some 0,11 kg per kg of condensate) at atmospheric pressure and 100 ᵒC. This mixture of condensate and flash steam still contains the sum total of the original condensate energy at 5 bar, namely 671 kJ/kg. And the total of this energy, when mixed in with make-up water at 25 ᵒC in the hot well, will cause its temperature to rise according to the formula: Energy = 4,19*kg*(T2-T1). One kg of condensate will thus increase the temperature of 2 kg of feed water by ((671+4,19*25)÷(4,19*2)) = 92,6 ᵒC.

We seem to run into a problem here, since water at atmospheric pressure boils at 100 ᵒC, so if we mix too much condensate with make-up water in the hot well, it will start boiling and excess heat will escape as flash steam through the vent. Theoretically (meaning no energy losses taken into consideration) at least it seems to indicate that with an open hot well there is a certain maximum percentage of condensate that can be returned before the hot well starts boiling.

Our attempt at calculating the maximum condensate returned, hinges on the following equation:

Energy of feed water at boiling point (100 ᵒC) = (Energy of condensate + Energy of make-up), remembering that 100% feed water consists of x% condensate and (100-x)% make-up water. Thus, to calculate the percentage (x) of condensate returned at which the water in the hot well starts boiling, the following formula applies:

100%*4,19 kJ/kg*100 ᵒC = [x%*671 kJ/kg]+[(100-x)%*4,19 kJ/kg*25 ᵒC].

Solving this equation for x provides us with a maximum condensate return of 55,5% before the water in the hot well starts boiling.

I am the first to admit that this is a purely theoretical exercise. Depending on condensate system heat losses, more condensate can be returned before boiling takes place. In fact, the more condensate is returned, the less make-up and water treatment will be required. But from an energy point of view alone:

  1. Incurring cost to recover all of the available steam condensate may not be the most economical solution.
  2. Aim at returning sufficient condensate to bring the hot well as close to boiling as possible, without causing feed pumps to cavitate. 

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.


Boiler Bits No. Title Category
1 Introduction: The Layman’s Guide to Coal Boiler Operation and Supervision Introduction
2 Requirements for a Boiler Control System Automation
3 What Level of Boiler Automation? Automation
4 The Mission of Steam Plant Operations Management
5 Principles of Combustion Technical
6 The Significance of Excess Air in the Quest for Optimized Combustion Efficiency
7 The Case for Oxygen Trim Control Automation
8 The Achilles Heel of Oxygen Trim Control Automation
9 The Nature and Magnitude of Energy Loss in Steam Boilers Efficiency
10 Expressing Efficiency of Steam Plant Efficiency
11 Smoke Control Principles Technical
12 The Influence of Steam Pressure on Boiler Efficiency Efficiency
13 Coal Boilers – How Light on Fuel? Efficiency
14 Coal Usage Based Efficiency Indicators Efficiency
15 Estimating the Temperature of the Fire Technical
16 Heat Transfer within Steam Boilers Technical
17 Firing Sunflower Husk in a Coal Boiler Technical
18 Optimum Condensate Return Percentage Technical
19 Coal as Determining Factor in the Combustion Equation Efficiency
20 Electricity Consumption of Boiler Fans vs. Flue Gas O2 Levels Efficiency
21 Adding Steam to the Hot Well: Anything to be Gained?    Efficiency
22 Boiler Water Level Control: Effects of Shrink and Swell Technical
23 Furnace Pressure and Induced Tramp Air Efficiency
24 Configuration of Draught Sensing Piping Technical
25 ID Fan or Stoker Leading Control for Boilers? Automation


Some of our clients are in the fortunate position to derive boiler fuel from their production process as a waste product. They are in the edible oil business, extracting sunflower oil from the seeds and being in the enviable position to transport the waste husk over a distance of less than 100 metres to the steam plant where it is fired in standard packaged coal fired boilers to produce enough steam for the whole plant.

But this benefit of free fuel does not come without challenges and one can really write volumes about it. One of the biggest challenges in my opinion is having to burn two widely differing fuels (coal and husk) in the same furnace and expecting both to deliver the same amount of steam at acceptable levels of efficiency.

Having been designed for coal firing the boilers perform well with coal fuel and combustion optimization is fairly easy to achieve by virtue of the fact that this is the default operating mode of the boiler. Firing husk with the same boiler is the culprit – it is foreign to the design of the boiler, but with a few make-shift arrangements (like adapting stoker sprocket sizes) users succeed to some degree to use it for raising steam.

Here are a few challenges I have experienced when firing husk (biomass) in a coal (fossil fuel) boiler:

  1. The control system must provide for two different control programmes with different settings, so that the operator can change over from coal to husk with the push of a button, without having to enter new combustion settings, except maybe for excess air requirements.
  2. I have found that in most instances where husk is fired the operator sets the control to manual mode and then sets the fuel bed deep and all motor speeds or dampers to maximum. Combustion air flow may be throttled to obtain a strong draught (low furnace pressure) to prevent the fire from burning back into the hopper, or at least, this is the general perception.
  3. Wasting energy with husk firing is costly – when the husk is finished, the user has to revert back to coal firing which adds significantly to the cost of steam production. Firing husk efficiently can extend the productive use of this free resource significantly.
  4. With too much combustion air and deep suction of the ID fan, burning fragments of husk is entrained in the flue gas and end up inside the grit collector where it can set alight unburned husk particles and cause a fire. I have seen this happen, resulting in a warped grit collector and shriveled flue gas instrumentation.
  5. Husk firing generates significant quantities of fly ash, which has a tendency to partly settle in the reversal chamber and partly in second and third pass tubes. Regular tube cleaning is required to maintain proper heat transfer. (We are currently investigating combustion settings for husk, as well as properties of fly ash, which cause it to settle inside tubes.)
  6. Because husk is high in volatiles (±70%) it ignites easily and burns out rapidly. The burning back of the fire into the fuel hopper is a constant threat and suitable measures must be applied to combat this phenomenon. Husk firing is best suited to applications where the steam demand is consistently high. The moment the stoker slows down as steam pressure increases the fire starts burning back into the hopper. Operator routines, such as de-ashing, can also create conditions favouring burn back. Apart from anything else let me say at least this: a deep draught has an inclination to promote burn back, rather than preventing it.
  7. A deep draught also causes significant ingress of tramp air into the combustion system. This is especially true with high volatile sunflower husk, and the effects of draught are clearly visible in the volume of combustion air required from the FD fan. With a deep negative furnace pressure combustion may be influenced stronger by ingress air than by combustion air supplied by the FD fan and control of the optimum air-fuel ratio. All the more reason to employ a control system designed for the combustion requirements of both coal and husk.
  8. Because husk burns out rapidly the rate at which it can be fed into the furnace is rather high – we have managed a linear grate speed of some 800 mm per minute without losing ignition or running a long fire. On the other hand the minimum required stoker speed with coal firing is some 15 mm per minute. This places a huge demand on the control system and boiler hardware to provide consistently stable combustion control when either coal or husk is fired.
  9. The risk of overheating stoker motors always exists where the boiler is fired with both coal and husk. And quite often the stoker motor does not develop enough torque to move the grate at minimum speed (when coal is fired).

This post was compiled by René le Roux for Le Roux Combustion, all rights reserved. Do you want to know more about efficiency of combustion or combustion optimization? Please contact us for your professional boiler automation, steam system efficiency and coal characterization needs.

Kindly note that our posts do not constitute professional advice and the comments, opinions and conclusions drawn from this post must be evaluated and implemented with discretion by our readers at their own risk.