STEAM ENGINES AND WATERWHEELS

[Stanley]

Boilers never suffered badly from internal corrosion under normal circumstances because the scale protected them. However they could suffer from 'Grooving' which was cracking of the steel plate on the inside of a sharp bend or at the interface of a connection with another rigid part of the structure. Basically this is because of stresses induced during expansion and contraction and I have always suspected that part of the reason for this breakdown of the structure of the metal was due to corrosion cells induced by the stresses.
The most common site for external corrosion was where the weight of the boiler rested on the seating blocks. This was always associated with damp boiler settings and could rapidly dangerously reduce the thickness of the plate. At the Ten Year inspection a selection of seating blocks were removed and the plate thickness tested either by drilling or Non Destructive Testing.

[Stanley]

A Lancashire boiler is a complex structure and over its life is subjected to enormous stress. Work out the internal surface in square inches and multiply this by the working pressure and you will get an amazing figure.... Constant stress like this was easily catered for by plate thickness and general design and over the years the construction became standard, the makers knew what worked from experience. The real problems arose from changes in pressure during their working life. Boilers in mills driven by engines where close attention was paid to maintaining constant pressure at all times could be remarkably long-lived. By contrast, a boiler in a factory where there was a wildly fluctuating load, such as a brewery or dye works where large quantities of steam were needed for heating vats suffered badly and showed it over the years.
There was another form of failure which could show up shortly after commissioning. The boiler had to 'get used' to accommodating the stresses and sometimes one component would fail as the boiler 'breathed' and settled into its work. The Bancroft boiler had a front gusset that was completely cracked but this was always accounted for by the surveyor as being "of long standing" and it was ignored because it was recognised that the boiler was still perfectly adequate.
When the very large walking drag-lines were built for opencast mining sites they had gussets in the corners of what could be a moving structure weighing hundreds of tons flexing as it moved over uneven ground. The designers made sure that the structure was overbuilt and expected some components to crack during initial commissioning. These were regarded as sacrificial and didn't affect the overall performance.

[Stanley]

The whole purpose of the boiler was to burn coal as efficiently as possible in the two furnace tubes and the firebeater was the man who made sure this happened. At the end of my time at Bancroft I had a wonderful firebeater, the late John Plummer. He had been firing boilers most of his life on everything from steam drifters fishing round Bear Island to Large banana boats from the West Indies. He knew his job and on this pic he's adjusting the variable speed gearbox on the Proctor Coking Stoker.
The art of burning coal is to use the minimum of coal at the right time and deliver it into the furnace in such a way that the over fire air and the draught through the firebars from underneath gives just the right combination to burn the smoke efficiently and keep a full bed of white hot coal right to the back of the bars with no gaps in the fire. In the early days this was done on fixed firebars by shovelling the coal in by hand through a door at the front that gave access to the top half of the furnace. This was skilled work and took many years to perfect. However, as time went on it was realised that investing in some sort of mechanised firing system could increase efficiency. The first stokers were crude affairs, coal was dropped on to a cast iron plate and thrown into the furnace by a revolving fan on to the fixed bars. If these stokers were carefully supervised and set just right they could give very good results but there was still the inefficiency of having to open the front door letting in cold air while the clinker was broken up and occasionally cleaned out. Our Proctor stokers went one better....

[Stanley]

The Proctor Wide Ram Coking Stoker was a good tool, there was one unit for each tube working independently. It had a small electric motor driving a Croft reduction gearbox which had two drives coming off it. One drove an oscillating ram in the box under the hopper which took coal regulated by a slide in the throat of the hopper and pushed it out on to heavy cast iron plates in the furnace front where it took its initial burn fed by air through holes in the low narrow door (You can see the glow in the image) as the volatiles (smoke) gassed off they were ignited by the white hot coals on the firebars.
The other output from the gearbox rotated a camshaft across the front of the firebox and this carried the cams that controlled the firebars.



The firebars were closely packed but with air gaps between them. You can see that they are raised at the back end to encourage the clinker and ash to build at this point before falling over into the ashpit immediately behind the bars and in front of the ash pit wall you can see further down the tube. The gab underneath the end of the bars was closed by a semi-circular door which could be raised using a chain from the front in order to get the clinker and ash out without opening the firebox door.
The front ends of the bars were hooked and each bar had one cam. The cams were set in a staggered fashion across the shaft and the effect was that as the camshaft rotated, individual bars were pushed forward about six inches until a point where they were all forward and the cams pulled the whole of the bed back ready to start the cycle again. This meant that any build up of ash and clinker at the back of the bars was encouraged to drop into the ashpit.
I think you can see that by regulating the coal feed, the speed of the drive and the draught the firebeater can control the depth and length of the fire to give the best burn commensurate with the demand on the boiler This worked well as long as there was an adequate load on the boiler. Where problems arose was when you only needed thin fires and you got too much underfire draught and leakage at the back of a short fire which cooled the fire down and resulted in incomplete burning of the volatiles, in other words, smoke!

[Stanley]

By the 1970s when I was in charge of the plant at Bancroft we were running into trouble with the 'Nuisance Man', the council inspector who amongst his other duties monitored the amount of 'dark smoke' emitted from chimneys. In earlier times mills were left alone, smoke was seen as a sign of useful work being done, in fact some owners encouraged their engineers to make smoke as a sign of how hard they were working and ho well the mill was doing. With the advent of the Clean Air Acts between 1956 and 1993 when domestic chimneys were included the mills were more closely monitored and regulated. It was almost impossible for us to comply because being on light load we had short fires and were getting too much cold air into the fire-bed causing dark smoke almost all the time.
I tried everything to cure the problem but in the end came to the conclusion that the only cure was to install under-fired stokers. I devised a scheme whereby we could do this at no cost and get perfect combustion. This involved burning our 300 tons of stock coal. The management hadn't realised we had this asset in the yard and told us to burn the stock, when it was gone they closed the mill using dark smoke as the excuse. So in effect I closed the mill.... I was not best pleased....

[Stanley]

Back to the boiler. I talk about balancing the draught against the fuel burning rate. This was done by using the dampers on the flue system. There were four main dampers (movable doors that could be adjusted to open or close flues. The two least used were on the economiser. One was a pivoted door in the flue from the main flue into the economisers. In normal conditions this was always open and the hot flue gases (in excess of 800F) bathed the nest of tubes in the economiser to pre-heat the boiler feed water. More about that later. The other damper associated with the economiser was the By-pass damper. This was in the main flue before the connies and was normally closed completely forcing all the gas into the connies. If the temperature of the feed water got too high it could be opened to allow the gases to go directly to the chimney.
This left the two side flue dampers which were at the point where the side flues entered the main flue behind the downtake. These were vertical slides controlled by wire ropes running over pulleys to counter weights at the front of the boiler where they were easily accessible by the firebeater. These were in use all the time the boiler was running and allowed fine adjustment of the draught so as to get exactly the right amount of air entering the furnace tube at the front.



A clean side flue at Bancroft in 1977. If you follow the left hand wall to the back of the boiler you can see the side flue completely closed blocking the flow of air in the flue. In normal running the firebeater had a fair idea of where his side flue dampers should be set and once his fires were settled he adjusted the flow by closing the dampers until a trace of smoke appeared at the chimney top and then opened them slightly until there was just a faint haze.



This mirror on the engine house porch was angled so that when the firebeater was standing next to the damper counterweights he could see the chimney top. Dead simple and effective!

[Stanley]

Really well equipped boilers had adequate instrumentation like draught gauges and CO meters but we had nothing, it was all done the old fashioned way!
One word about draught, it's a widely held misconception that the chimney produces suction to draw air through the furnaces. In a way this is true but the mechanism is that because the gas column in the chimney is hotter than the outside atmosphere it is lighter and produces a pressure differential in the flue system. At the boiler front the atmospheric pressure in the furnace is less than ambient so the atmosphere forces air in through the boiler front. This explains why a hot fire has more draught, a flue draws better in cold weather and the higher the atmospheric pressure the more draught you have. This was why the boiler house doors were always wide open so as not to impede the air getting in. Paradoxically this meant that the boiler front was the coldest place in the mill in frosty weather.

[Stanley]

A most important part of the plant was the economiser. At Bancroft we had a Green's from Wakefield exactly like this one in the image. You can see how it worked, a nest of cast iron tubes was bathed in hot flue gas and pre-heated the feed water for the boiler. It's important to realise that in a perfect installation the water was forced by the feed pump into the boiler via the economiser and so it was at high pressure and could accept 'superheat' above boiling point at atmospheric pressure. The accepted theory was that every ten degrees you raised feed water temperature you increased overall thermal efficiency of the boiler by 1% Put another way, ten degrees rise saved one ton of coal in every hundred. When you consider that it was quite possible to raise feed water temperature by over 300F you can see the saving.
Note the scraper gear fitted on top of the economiser. This raised and lowered cast iron scrapers on each tube which scraped off the worst of any flue dust or unburnt coal off the tubes increasing their conductivity. As a matter of interest, the flue dust that gathered in the bottom of the connie was always black as unburnt hydrocarbons in smoke were the most likely to be deposited on the tubes.

[Stanley]

Economisers were classed as a pressure vessel by the insurance companies because they were working at above boiler pressure. They had to be fitted with safety valves and were inspected annually at the same time as the boiler. The main factor the surveyors were interested in was the wall thickness of the CI tubes which were eroded over the years by the flue gas and worn by the scrapers sliding up and down them. Bancroft's connie, like many that age was eventually condemned. The cure was either a new economiser or to operate it at atmospheric pressure. This meant that the maximum temperature that it could be run at was less than 100C. The direct connection to the boiler had to be discarded and what happened at Bancroft was that Brown and Pickles converted it to a feed water heater. Feed water was pumped through the tubes and returned to a 'hot box' in the cellar from where it was pumped by a separate pump into the boiler. For many years it worked like this, the Weir steam pump sent the water round the connies and the Pearn Three Throw pump took it from the hot box and forced it into the boiler. Though not as efficient as the original plant it was still a considerable saving and avoided the capital expense of a new connie.



The hot well box in Bancroft cellar. The float switch controlled the Pearn three throw feed pump which picked the water up and sent it to the boiler. The rate of feed was adjusted by moderating the flow from the Weir pump. This was not an exact science!

[Stanley]

During 1976 at Bancroft we were in a bad way. The Pearn three throw pump was so inefficient we were having to pump water direct to the boiler without pre-heating it. During the day, despite this we were losing water all day. At teatime I had to put the Weir pump on, go home for my tea and come back an hour later to stop it when we had a full glass of water. Our coal figures were terrible and something had to be done. I persuaded the management to let me spend £400 on a big refurbished Brown and Pickles three throw pump and the electrical wiring and John Plummer and I carried about three tons of wet concrete into the cellar in buckets to pour a new bed. We installed the pump...



John suggested that instead of regulating the water to the boiler by controlling the Pearn pump we ran it all the time and fitted a bypass pipe with a valve in with which to control the output. I had reground the valves on the Pearn and while it was not perfect it was a lot better. John's idea was brilliant and worked perfectly. The B&P pump was far too big and so I ran it slowly. Because it was slow it could pump water almost at boiling point so we could extract the maximum out of the connie and the saving was about 3 tons of coal a week. John could control his water level with ease and so could work the boiler far more efficiently. Besides that we were safer and still had the option of the Weir in an emergency. Probably the most effective single improvement I ever made at Bancroft. The float switch on the hot box controlled the B&P pump and the sound effects from the cellar changed completely. We had the constant ringing of the gear drive on the Pearn and the intermittent low growl of the B&P pump as it tripped in, grabbed the hot water and forced it into the boiler. Very comforting.....

[Stanley]

I said above that the big pump was Brown and Pickles, this is wrong, it was made by Henry Brown and sons, the precursors to B&P. This means it was pre-1929. Jim Fort could remember that pump as he had turned the crankshaft out of a solid billet of 70 ton steel. He said it took about four days work to make and he remembered it because he had made a mistake which had to be rectified by modifying the end brass.



A brand new pump at Henry Brown and Sons, Havre Park. In the 1920s.

[Stanley]

The cellar at Bancroft looking down the high pressure side of the engine towards the Electricity distribution board, the exit and the back end of the engine where the air pump was on the LP side.
The boarded over pit in the floor is the old hot well fed via a ball valve from the overflow from the air pump coming down the right hand side. Originally the condensate returns from the mill came back to here but when the connie feeds were altered after they were de-pressurised, these were re-piped to the new hot well tank. The various drains from the engine still came back into here apart from the main cylinder drain cocks left open on starting, they always discharged to atmosphere via a small well on the outside of the engine house so as to make sure they always had free passage.
This arrangement meant that as the Pearn picked up the feed water from this old well we were usually feeding water from the lodge. The only exception to this was that if the lodge water was badly discoloured by silt during heavy rainfall I could turn mains water into the well to keep the feed cleaner. The lodge water was, if clear, good feed water and easily treated with the chemicals from Charlie Southwell. One point to note is that I had it set so that the well was always overflowing slightly back to the drain. This meant that any oill carried over from the engine floated to the top and was drained away before it could get into the feed water system.
The Weir pump picked up from this well and when it was running the exhaust went into the water minimising the heat loss. If necessary, say when filling the boiler after shut down, I could alter the valves and the B&P pump drew straight from the well with the mains water turned full on. When the mill was first built the boiler was filled by a separate supply through a 3" pipe laid across the field into a small catchment trap in Gillians Beck well above the mill so that gravity did the trick. This pipe was completely blocked by the time I took over and so was never used. One small point which I didn't discover until the mill closed and we were shutting off all the water mains to the mill. The sprinler system was fed direct from the main before it was metered. This was standard practice. What wasn't standard was that the mains water supply to the old hot well was opicked up off the sprinkler main and so was not metered. Definitely a bit of sharp practice!

[Stanley]

Cutting the water supply off after closure in December 1978. There was a very good head of water on the main coming down from Whitemoor and so Bancroft didn't need a static water tank for the sprinkler system, the main supplied plenty of pressure. One regular weekly job was to test the sprinkler system. To do this you opened a relief valve in the rising main and this triggered the water turbine in the feed which drove the sprinkler bell on the outside of the wall. Once you heard the bell ring you could shut the valve and the bell stopped. I can still hear the distinctive sound of that bell as I write this.....
The sprinkler system at Bancroft was a wet system as opposed to some installations which were dry until a sprinkler head was triggered. This meant that in an unheated mill they were prone to frost damage, so the last job was to drain the system after the main was closed. Shutting down a mill is a miserable time......

[Stanley]

An important but often neglected part of the boiler and engine was the steam main from the boiler to the engine. Many mistakes were made in the initial design of the plant. Ideally the main should be as short as possible and always be inclined upwards so that any condensate in the pipe constantly drained back to the boiler, the last thing you want in a cylinder is 'carry over' from the boiler. The length was important as there is no such thing as a perfectly insulated pipe and to superheated water (which is what steam actually is) the walls are cold and the longer the main, the greater the amount of heat loss and condensation. This was made worse by the practice of not insulating the flanged joints on the main. I think the idea was so that you could immediately see any leaks but in practice you would soon know about them anyway. Almost all these joints leaked to some extent when starting from cold but tightened up as the pipe expanded.
At Bancroft the original plan was to site the boiler much closer to the engine but when they were excavating for the build they hit an anomaly that they didn't fully understand and moved the boiler further away to avoid it. Eventually when the mill was being demolished we found that this anomaly was a mineral deposit that formed and grew over the upward seepage from a deep bed of peat. We knew about it and called it the Bancroft mushroom because it raised the floor in the end of the warehouse and the tape room above. We had to adjust the level of the tape sizing machines with blocks under the feet from time to time.
Luckily the pipefitters used their heads and gave the main plenty of fall back to the boiler and we were always careful about water in the pipe on starting. At Ellenroad the main was very long because of the size of the range of boilers, five in all, and a very large Hopkinson separator was fitted in the line just before the steam entered the engine. This could be blown down on start up and had a large steam trap constantly discharging to atmosphere outside the house ensuring that there was no water build up. Note 'to atmosphere', it was essential that all drains from the engine had perfectly free egress, I have seen mistakes made by attempting to capture this hot condensate in the general condensate main back to the hot well. This meant that there was back pressure in the drain and it couldn't work efficiently.
All this is quite boring I know but absolutely essential. The consequences of a large slug of water getting into a cylinder could be serious.



It was water in the low pressure cylinder that caused this smash at Crow Nest Mill in Barlick. Well worth avoiding!

[chinatyke]

Why couldn't pressure relief valves be fitted to the cylinder end to void any water before it caused hydraulic fractures as the pressure under these circumstances would have been extremely high? Surely PRVs come in all sizes and pressures?

[Stanley]

They were fitted P.



You can see them on this pic of the back of the HP cylinder at Bancroft. They are the brass cylinders mounted above the indicator cocks at each end. The LP had two as well and there was one fitted on the large connecting pipe in the cellar between the two cylinders. The problem with the ones on the cylinders was two fold; unless you inspected them frequently to make sure they were free they tended to get glued down by metal polish running into the seat through the escape port, this could glue them down solid. The other problem was that if you got a serious slug of water there is so much volume so quickly that in order to cope with the flow the valve would have to be enormous. They were usually set to open at 10/20psi above expected maximum pressure which could be higher than normal steam pressure in an HP cylinder working with a lot of compression at the end of the stroke. A further problem on LP cylinders was that higher pressures could be expected when starting the engine and so these were set at well above the normal working pressure. The consequence was that a really bad slug could still cause damage. At Ellenroad, running under no load conditions for demonstration water could gather in the LP cylinders, for some reason this was always worse in the left hand engine. This was more a shock at each end of the stroke than a simple increase in pressure and never shifted the PRV even though it was free and set at a low value. I used to open the drain at the front end of the cylinder to relieve the problem. To give you an idea, even with this small amount of water you could hear the thump and feel the movement in the cylinder cover when it occurred.
So, the problem of a slug could only really be addressed by making sure they never happened. Watching boiler water levels, keeping the boiler water in good condition and ensuring that the rubbers on the valves in the air pump were in good condition.

[chinatyke]

Thanks Stanley. Regardless of the pressures that were expected in the cylinders during normal operations including start ups, surely these PRVs could be set to operate at much higher pressures as they were protecting the cylinder from failure during hydraulic events? What I mean is the cylinder itself could be regarded as a pressure vessel which I assume would catastrophically fail at much higher than working pressures, and so the PRV could be a very large dump valve operating at well over normal working pressure. In effect the whole end of the cylinder could be a PRV, or couldn't it?

I had a wonderful experience at the chemical works when a 30bar PRV lifted and didn't re-seat and vented an entire compressed gas storage vessel over the next 2 hours, very exciting and frightening.

[Stanley]

I know the feeling! We did this once a year when we blew the boiler off in order to empty it and cool it down ready for the insurance surveyor and the fluers coming in the next day. I did it by lifting the arm of the low water valve and lodging it on a couple of bricks.



10 bar escaping through a 3" pipe was impressive and of course carried on much longer as the water was producing more steam as the pressure dropped and it flashed off. One woman told me that her cat usually vanished for a couple of days when we did it!
As for bigger PRV on cylinders. Really big events like a massive slug were such rare events that it wasn't considered a big enough risk to be covered. My valves at Bancroft were free, I made sure of that, and in six years I never even got near lifting them because we looked after the plant. There is also another consideration. The engine manufacturers didn't suffer if a cylinder was smashed as they would be called on to replace it. This isn't far-fetched, when the High and Intermediate cylinders were replaced on the original triple expansion layout at Ellenroad to convert it to tandem and get more power, very good lubrication was installed on the new HP cylinders and none at all on the original LPs. It wasn't their problem! The event at Crow Nest was a good earner for Brown and Pickles.....

[chinatyke]

Looking at your picture of the PRV reminds me of when we (complete novices and not engineers) installed a small vertical package boiler and called in the insurance inspector to commission and certify the installation. His immediate concern was that we had fitted a short 90 degree bend on the relief line and not used a swept bend. We were made to remove the fitting before he would proceed, and this left the PRV vent temporarily pointing directly towards an interior wall during commissioning. As the engineer was adjusting the PRV to set it above the working pressure, the PRV lifted and because he'd backed off the locking nut the main nut ran up the thread and the boiler blew down to atmospheric pressure. Never seen 3 people leave a room as quick as it cleared us out. Cleaned the walls though!

[Stanley]

You've reminded me of something my dad told me once. After WW1 he was working at Armstrong's at Trafford Park and they did some experiments comparing swept bends with 90degree bends. He said that they found that in some cases the 90degree bends were more efficient as some designs of swept bend induced turbulence and slowed down the flow. We were having the conversation because I had noticed that I got more pin hole leakage on swept bends in heating lines in the shed than on elbows. he said it could be the turbulence that was causing pressure induced corrosion cells but that the elbows were thicker metal.
I like little conundrums like that.....

[chinatyke]

Were the pinholes always on the outer edge of the swept bend where the steam was moving faster? Was it just the velocity of the gas coupled with thinner walls? Are swept bends made by a different process such as drawing or bending as opposed to casting of elbows and would this lead to thinner walls on the outer edge? Just thoughts...

[Stanley]

All those are possibler factors but their research was purely on flow rates. Another factor in long heating runs was water hammer when starting from cold.....

[Stanley]

When first installed the boiler at Bancroft had a Hopkinson condensate return system, a very large specialised trap made so that when full it could discharge water to the boiler even though it was under pressure. For some reason this had not been used for years, all condensate from the heating mains was returned to the hot box in the cellar. In many mills this valuable feed water was allowed to run to waste and after WW2 B&P had a good sideline installing these systems in mills. The sales ploy Johnny used was always the same, he'd put it in for nothing if the owners would sign up to paying him a high proportion of the fuel savings for five years. This almost always worked and they did a lot of mills.
On a similar subject many visitors to the mill used to ask me where the steam was. They were used to seeing locos in films blowing steam out of every possible orifice and expected every steam engine to do this. We didn't have any and I always told them the reason why, steam escaping is waste and we kept it in the plant where it was doing most good.....

[Stanley]

One of the unavoidable py products of running a coal fired boiler was the ash, flue dust and clinker. We kept the flue dust separate.... but the clinker and ash were tipped in a pile in the corner of the yard. The funny thing is that we never had to get anyone in to take them away, everyone knew that it could be had for free at Bancroft and the council often came for a load. Individuals came as well, it was popular for paths in allotments.... Here's my old doctor Arthur Morrison bagging clinker to take to Thornton.

[Stanley]

We always kept flue dust separate because it was the favoured base used by flaggers when laying stone flags. Because of its sharp edges and fine texture it didn't creep like the rounded edges on river sand when flags were laid on it. There was always a demand for it right up to when the mill closed. Here's my old mate Jack Platt wheeling some away. He was putting flags down in his back yard and knew the advantages of flue dust.