Boiler Feed Pump

[Alan Worland 1Participant @alanworland1]

[Alan Worland 1Participant @alanworland1]

I have read Model Engineers Workshop for a long time but only just joined the ‘forum’! and plant my first post!

I was recently having a rumage in the loft when I thought I would get my old steam boat out for a play. Built by my Dads uncle about 1950 I think.

It consists of an aluminium and ply hull about 4 foot long, powered by a paraffin fired copper water tube boiler – 5 inch diameter by 9 inches long with a 2 1/2 flue fitted with 8 water tubes and a superheater.

This feeds a Stuart Turner Sun engine and drives a single prop.

I can remember my Dad having this afloat at Victoria Park boating lake in London when I was a wee nipper. It used to go well but lacked any sort of control – set the rudder, open the steam valve and hope you, or some helpfull person would ‘catch’ it round the other side! all very nerve racking! Of course then the hand pump would be used frantically to replenish the boiler!

So, my current project consists of equiping the boiler with some sort of gas burner to enable more control (and perhaps a bit quieter), re mounting the engine horizontal and building a gear driven drive for a boiler feed pump I have, and here is my problem, I can’t seem to find out what sort of water volume it should be pumping to maintain boiler water level. The Sun engine is a single acting twin of 3/4 bore and stroke.

Is there any tables which give a guide to steam (water) consumption for a given engine size? I realise that this would probably only be a guide, but a guide is better than not a clue!

[Alan Worland 1Participant @alanworland1]

I think I am getting somewhere!

I have found out that water when turned into steam will expand about 1600 times – so 1cc of water when boiled will produce about 1600cc of steam.

I know my engine has two cylinders of 3/4 inch bore and stroke so I can determine the stroke of my feed pump when running via a worm drive (36:1) off the engine.

I will be allowing some sort of ‘safety margin’ so the pump will deliver more than I believe the engine is using, the excess could go to waste?

My question is, what sort of safety margin shall I allow, 50% seems a good starting point but it would be nice to know if this is considered ‘about right’

[Les Jones 1Participant @lesjones1]

Hi Alan,

                  I have had nothing to do with building or designing steam engines but there may be a flaw in your reasoning. Is this expansion factor for steam at atmospheric pressure or at the working pressure of your engine. If the figure is at atmospheric pressure which is about 15 PSI then if the engine runs at 90 psi above atmospheric this is 105 psi absolute. so the volume of steam would only be 15/105 (about 0.143) of the volume at atmospheric pressure.  I would wait for someone with more knowledge to reply before finalizing your design.

Les.

[Alan Worland 1Participant @alanworland1]

Hi Les, Thanks for your input!

Your’e right this is at atmospheric pressure – my reasoning was that the water going into the boiler starts off at atmospheric and the final steam out of the exhaust is at atmospheric pressure?

But I think your right – I need some guidance!

[Billy MillsParticipant @billymills]

Hi Gentlemen,

How about condensing the steam, collecting in a small tank then returning the water to the boiler when the tank level reaches a set amount?  The recycling might not be 100% but you are not crossing the Atlantic.

 

regards,

Alan

[mgjParticipant @mgj]

Think of the steam in terms of mass flow rather than volume. ie you turn a CC of water into steam. So you now have 1600cc of steam at 1 bar, but its still only 1cc of water – =1gram. You have 800c of steam at 2 bar, but its still only 1cc of water =1 gram.

 

Its energy potential and the energy input to get it to 2 bar is greater, but you still only have a gram of steam = 1cc of water.

 

Also the limiting factor will not be how much steam the engine uses, but how much steam the boiler can produce, because if the engine “calls” for more than the boiler can make, boiler pressure will drop. I think it would be quite easy to do the maffs for that. You can work out volume quite easily. Thermal input is not so easy, so crank the fire up and measure the drop in a given period, and then you know how much water you are using.

 

You will need to tether the boat in water to give you a prop load of course. One of htose model aircraft optical tachos and a couple of dabs of paint will give you shaft  revs and hence a gear ratio.

Edited By meyrick griffith-jones on 17/12/2009 23:31:00

[Alan Worland 1Participant @alanworland1]

Alan, I must admit I had thought about condensing the exhaust for return, but at this stage of the project I was just trying to get the basics right (or about right!) to get it running again with a feed pump capable of delivering the correct(ish) amount of water.

Are there figures relating to gas burner output and steam production? I realise this would depend on boiler efficiency.

[Billy MillsParticipant @billymills]

Gentlemen,

You don’t need to do the thermodynamics to top up the boiler- which seems to be the original concern. Simply that if you have lost 50mL of water from the boiler -49 mL of which is now condensate- you can return 49mL back to the boiler.

 

You only need a Bang Bang controller to do this, once you have a little condensate run the pump for a short pre-set time to return SOME (but not all)  of the water to the boiler then wait till the level gets to the trip level before the next run. Simple.It adjusts itself, no complex algorithms needed! But it might not be what you need.

 

I appreciate that you are looking at some kind of advice on consumption rate however  small heat engines will suffer from a very large number of losses from many different variables, any estimates may be a very long way out. Superheating greatly complicates matters, it puts in a lot of additional energy without increasing water consumption. I would suggest that after 50 years of non-use the boiler should be pressure tested then you can go for a test steam and quickly find out the magnitude of the problem. You may not have long enough trips to require boiler topping at all!   If boiler water level went down too quickly all that time ago then steam loss might have be a problem or excessive heating and venting. It would be prudent to check these possibilities before working on a solution to another problem that may not exist.

 

Another kind of steamer was at the Sandown MEX last weekend,  a Stanley Car. Did a quick Google and found a Stanley website, one model -not the one at Sandown- had two 4″ pistons and a boiler max pressure of around 550psi, about 25HP when trundling along but 125HP before the pressure dropped. That was semi-flash steaming!

 

regards,

Alan.

 

 

 

 

 

 

[mgjParticipant @mgj]

Alan thank you about the advice about not doing the thermodynamics.

 

I think that was my point – the only addition was to do an empirical test under load so one found out how much water to pump before rather than after the event.

 

Which is usually a wise approach.

Either way, condensing could be a little awkward? Space, for one, and how much surface area? Back to a spot of maffs to work out flows (which is where we came in) and rates and joules and heat transfers and surface aeeas.? Unless one wants to make several.

It is not difficult maths in any way, but you need a mass flow rate – (and another (vacuum) pump probably?).

[Martin CottrellParticipant @martincottrell21329]

Hi Alan,

 

A few years ago now, I attended an open day at Cheddar Models who specialised in the design and production of small steam plants for model boats. They had a small gas fired steam plant running unattended on display, which kept the boiler water level topped up automatically by a mixture of electronic & mechanical control.

 

I can remember that there was an optical electronic sensor placed on the water level sight glass which switched the pumping mechanism on and off according to the water level. I can’t remember whether boiler filling was done by injector or pump although the simplest ( more reliable?) system would seem to be a pump with a two way valve, servo operated & switched by the optical sensor, which would either send flow to the boiler or return it to the water storage tank as necessary. I seem to recall that there was also a pressure sensor incorporated which allowed control of the burner flame to avoid excessive steam blow-off when the engine was running slowly or stopped.

 

I believe that Cheddar Models are no longer trading although I think Stuart Models may have taken over their range. In any event, I would have thought that a chat with your local model boat club would more than likely prove fruitful.

 

Regards, Martin.

Edited By Martin Cottrell on 18/12/2009 10:16:20

[Billy MillsParticipant @billymills]

Gentlemen,

There is an almost infinite cold sink in the form of the “Lake” a few cm from the boiler so a condenser should not be tooooo difficult! However the real unknown is what the original problem is. So we are all speculating away-as we do- without any hard data.  

 

It would be very good to know if the boiler leaked, a  water pressure test would be a great help. Leaks elsewhere are very possible as is the issue of over firing. A running test- after the boiler safety test- is what we need.

 

Martin’s memories are interesting. The idea of limiting the fuel flow by boiler pressure is not quite new but very sound, that’s what stokers did by the Armstrong method! 

 

Boiler topping by water level is a tad difficult as the sensor has to see a clear liquid, a simple beam break sensor would only work if dye was added to the boiler or a float fitted to the tube. However you could use the glass tube as a cylindrical lens with a dramatic change of focal length when clear water was inside.

 

You can invent servo systems to do almost anything useful /useless but they always introduce other issues that need to be thought about.  For example, water level sensing in a stationary model is relatively easy but when moving about can become unstable.

So you put in a small leak to damp the level indicator, it furs up when someone uses tap water and the boiler runs dry.

 

Yes, there is enough material here to keep us all speculating away well into 2010!

 

regards,

Alan

 

 

 

 

 

 

 

 

 

[CirclipParticipant @circlip]

After all the Maffs have been applied and the hypothesees? stated, where in the equasion do you apply the third law of Murphy???

 

  The optical sensor system is still being investigated and requested by lots on the toy boat sites and the ultimate steaming system, Flash, is being deeply investigated by a guy that knows his onions and is applying PIC’s to control systems for both boiler and Gas feed. In NO way am I criticising his application cos without guys like this we wouldn’t have had any of the old”Masters” of experimentation, BUT, does the average player with toys need to have such sophistication?? Do we really need an IT qualification to enjoy a quiet afternoons playing??

 

  With regards to feeding the beast, whats wrong with the simple engine driven feedwater pump with a simple bypass fitting?? Worked successfully for years before chips with everything? With the Sun being a thirsty engine with plenty of power, an engine driven pump won’t sap much power away, but with some of the smaller engines is anything more than a hand operated feed pump required??

 

  When you go for a “Session” with one of your toys, set a stopwatch going when you start and time to when you stop to have a “Rest”. Many would be surprised that what seems ages in fact is far less than they thought so the added complication of some control systems doesn’t justify the added complication or frustration.

 

  Regards  Ian.

 

Forgot to add, if the “Roarer” burner system ain’t broke, don’t fix it wiv Gas.

Edited By Circlip on 18/12/2009 12:57:31

[Alan Worland 1Participant @alanworland1]

Wow! That a lot to take in, thanks to all.

This is where I am at::

I have discovered (from Stuart) that the engine should be mounted horizontally to enable the sump to retain oil for splash lubrication. Doing this has moved it forward into the space where the paraffin burner tank was located.

As I remember this burner was an all or nothing device, which was great if you required flat out power, but I want more control over the flame with a requirement for more sedate running.

Work is progressing on the feed pump drive which is going to use a worm on the crank (36:1)  driving a 1/4 in bore pump with the stroke to be determined.

Probably over the Christmas period I shall fire it up in my test facility (bath!) flame is going to be supplied by  my range of Seivert burners and I hope to establish the maximum size (kW) of burner required for full power to maintain boiler pressure.

While this is going on I shall be counting pump strokes from the hand pump (3/8 in bore) mounted in the hull – hopefully I can calculate the water consumption!

I think prior to this it might be prudent to carry out a hydraulic test of the boiler.

I am not sure if this is to an acknowledged design, it appears well made from 3/32 seamless copper both for the outer and inner shell with diagonal cross tubes all siver soldered together.

The Bassette Lowke gauge fitted reads full scale 60 psi? which would suggest a working pressure of 40? (seems low) any thoughts on test pressure?

I am trying to basically finish off the boat to a usable level and am trying to keep things simple!

[mgjParticipant @mgj]

Well I rather agree – even with a dirty great big heat sink only a few Cm from the system, you still have to have enough exchange area available to shift heat at a rate, and know how much of that heat sink needs to go through the condenser as a minimum – and all and all.

 And it has to be maintained at a  lower pressure than the exhaust.

(Rather than dye, I see a whole capacitor belt gauge being fitted, or you could use a strain gauge to tell you how full the boiler was.) 

 

Alternatively, if you have a mean rate of consumption, you can just turn on the tap for a couple of mintues (and you know roughtly how much the pump will supply). Most modern RC sets have timers have  settable alarms. It may not be perfect, but it will work just fine

[Alan Worland 1Participant @alanworland1]

Well, I haven’t had it all running yet!

During the hydraulic test I had a weepy safety valve (I took the spring out and fiited a collet instead) gound it in with ‘T Cut’ seemed ok. Also had a weepy gauge glass, npped it up and the end broke off the glass! I think the packing had lost all its ‘softness’ fitted O rings it seems fine now.

I drained the boiler down to a working level and then fired it up on the bench for a steam test and to blow out the engine supply pipe, which was disconnected at the engine end. I used a large burner on my Sievert and had it just ‘ticking over’ up the flue and after a short while up came the gauge to about 50 psi with a hovering safety valve a good blast from the steam valve cleared the supply pipe – not sure if there was any dirt within but it seemed a good idea!

Operating the gauge glass blowdown valve resulted in steam coming out but very little water and the water level took a long while to rise again, I decided that the bottom fitting must be blocked. When it had all cooled down I removed the new glass to unscrew the bottom fitting – and yes you guessed, it sheared off flush!

I managed to remove the sheared off bit by gradually taking its diameter out in stages until I could ‘rake out’ the remains (tapped O BA and totally solid)

I removed the top fitting to see how clear that was – ok, and noticed that there appeared to be soft solder in the thread?  used as a sealant? the top fitting was very tight with the thread totally covered in solder (tinned?) never heard of this before.

I shall reconstruct the bottom fitting and continue!

[mgjParticipant @mgj]

Well I am on your side with sympathy, even if God is not -yet.

[SteamshyParticipant @steamshy]

The soft solder may have been used as a compression sealant around your guage glass, have used it successfully myself, just thin solder coil wound seems ok. Try the guage glass drill when ever possible, Shut steam, shut water, open drain, open steam cock slowly and blow clear,open water cock and repeat.Close drain valve, the water level should return smartly to the true level if drains were blown clear

[Author]

Posts