gas pressure and consumption

[theimi]

Hi guys,

 

wasn't there for a long time, hope everybody is fine

I'm running a blower operated gas(lpg) forge. I've got a pressure regulator at the tank and I use a needle valve and an air regulator to control my flame (in difference to most other folks I do not use a valve for the air but regulate the speed of my blower which works really nice for me.)

I'm busy building an electronic temperature control for my forges and what keeps me awake at night is the following problem:

If I would like to have a really low temperature I have two possiblities: 1) I crank down the pressure regulator and adjust the air to get the right flame or 2) I reduce the amount of lpg by using the needle valve and again adjust the air to get the right flame.

Is there any difference in gas consumption or any other disadvantage while using the second possibility? What's the difference in general?

For my electronic control it would be much easier to keep the pressure regulator on max. Haven't found an electronically operated pressure regulator til now, at least in South Africa....

Sorry, if my explantions are confusing you. I may chat to you with no problem for hours but explaining technical problems shows me that english is not my mother language :-)

 

Frank

[Buzzkill]

For this purpose the regulator and the needle valve accomplish the same basic function; they restrict the volume of propane supplied to your burner.  Since temperature is determined by the amount of propane burned at the right air mixture, both methods should give you the same fuel consumption for a given temperature over time.

Typically this type of setup will have an idle pathway for both the air and the fuel which is always open when the forge is in operation and set for about the lowest stable flame possible.  In your case you only need one pathway for air if you are able to control the blower speed accurately enough to work at both the idle and high settings.  A second fuel pathway with a normally closed solenoid would be opened by your temperature control system, and that same signal should be set to increase your blower output simultaneously.  Once you have both the idle and high fuel to air ratios set properly then the PID should cycle the high circuit on and off as needed to keep the forge within a set temperature range.  I would run the regulator at a high setting and use the needle valves to set the fuel flow on both the idle and high output fuel circuits.

On a safety note, at a minimum you should have a normally closed solenoid in your fuel supply line close to the source, using the same power circuit to open it as powers your blower, so if your blower ever loses power, all fuel going to the forge is immediately cut off.  Once a forge is up to temperature there is little chance of a flameout, but there are sensors and circuits that can be used to cut the fuel if a flame is not detected as well.

[WayneCoeArtistBlacksmith]

Actually, the regulator controls pressure, a needle valve controls volume.  For my Ribbon Burner I set my regulator at about one half pound of pressure and use the needle valve to control the volume of gas.  I use a 2" gate valve to control the volume of air.  When I used a venturi burner forge I ran the pressure at the regulator from five to seven pounds of pressure.

Let me know if I can help you,

Wayne

 

[Buzzkill]

Yes, Wayne, but in this context it accomplishes the same thing.  You could run your ribbon burner with no regulator and use just the needle valve or vice versa.  In simple terms a regulator is just a valve that uses a spring, a diaphragm, and gas pressure, rather than a handle, to determine how much the valve opens.

Without knowing how large the gas inlet orifice for your ribbon burner is compared to what it was in your venturi burner, no determination on which uses more volume of propane can be made.   One half psi running through a 10 foot diameter pipe is WAY more volume than 500 psi through a 0.025 inch jet for example.

[WayneCoeArtistBlacksmith]

Well, not exactly.  It accomplishes similar results.

6 hours ago, Buzzkill said:

Yes, Wayne, but in this context it accomplishes the same thing. 

 

[Buzzkill]

Not trying to nitpick here, but it is exactly the same.  The pressure at which propane enters the mixing tube of a blown burner is for all practical purposes irrelevant.  The only thing that matters is the volume of gas. No air needs to be induced in a forced air system, so whether you have a 1/2" fuel delivery line at very low pressure or a 0.023" mig tip at much higher pressure, the volume of propane must be the same in either case in order to get the heat you want.  I don't recommend it, but you could run the fuel line directly from the propane tank and use only a needle valve to control the volume of fuel.  The pressure after the needle valve would be way lower than the tank pressure just like it is with a regulator.  The way you have yours set up is the preferred way.  It's best not to have full pressure fuel lines anywhere near the forge, and with a gauge you get a visual indicator which helps to easily reproduce preferred settings in the future.  The needle valve in the system is really just fine tuning and may help smooth out any pressure pulses from the regulator as the valve inside it opens more or closes more to keep the pressure downstream close to the specified pressure.   Both the regulator and the needle valve affect the volume of propane that is flowing to the burner. Both of them will also affect the pressure downstream from them in the system.  Again, a regulator is a valve.  It just happens to have somewhat automatic features built into it for opening and closing as needed to get the desired result.

I wouldn't have bothered responding, but it seems frequently there is confusion about this topic. People ask what pressure they should run at - which I understand.  We don't have volume meters on our forges usually, but we typically do have gauges that show the psi.  The problem is that what seems like a relatively small difference in systems can cause them to run at significantly different pressures, but the volume of propane needed (at the right air mixture), especially for a blown burner, is really what is important for heat output.  In NA systems since the fuel delivery induces the air and mixes the fuel and air at the same time, the diameter of the orifice and the pressure at the orifice become much more important.

[theimi]

Thanks for all the information :-)

I use a temperature control as described by Buzzkill in his first reply for a long time already (idle pathway and heating pathway) and it works perfectly for all normal forging operations. The disadvantage is, as soon as it reaches the specified temp the PID switches into idle mode and the temperature goes down instantly for at least 50...100 deg celsius, the PID switches into heating immediately again, and so on

But now I would like to have something with a more stable temp for different special uses (making mokume for example). It should not be too difficult by using an arduino or similiar PIC to program a nice regulating system, which controls the heating on a much finer degree, for example the more the temp approximates the specified temp the less power (lpg/air) the burner needs. Hope my explanations are intelligible

Thanks again, 

[Andy98]

On 9/1/2016 at 3:43 PM, theimi said:

Thanks for all the information :-)

I use a temperature control as described by Buzzkill in his first reply for a long time already (idle pathway and heating pathway) and it works perfectly for all normal forging operations. The disadvantage is, as soon as it reaches the specified temp the PID switches into idle mode and the temperature goes down instantly for at least 50...100 deg celsius, the PID switches into heating immediately again, and so on

I've never used a PID controller for a forge, but I have used a lot of PID controllers and done a lot of other control applications. Disclaimer aside, are you sure you're actually doing PID control? Your description sounds more like on/off control. If it is actually a PID controller, and it has digital control outputs (and not just digital alarm outputs) then it should be able to do PWM (pulse width modulation) to vary the output and achieve something close to "in-between" values - see PWM here.

If it's doing PWM and you're not getting good control then maybe you haven't configured it correctly? It could be poorly tuned, or have too long of a duty cycle, etc..

FWIW, in process industries at least, you wouldn't typically have an "electronic regulator" - your controller becomes the regulator, and you wouldn't have it controlling for pressure - you'd have it controlling for temperature. It would naturally adjust for input pressure changes as the tank empties.

[theimi]

11 hours ago, Andy98 said:

I've never used a PID controller for a forge, but I have used a lot of PID controllers and done a lot of other control applications. Disclaimer aside, are you sure you're actually doing PID control? Your description sounds more like on/off control. If it is actually a PID controller, and it has digital control outputs (and not just digital alarm outputs) then it should be able to do PWM (pulse width modulation) to vary the output and achieve something close to "in-between" values - see PWM here.

If it's doing PWM and you're not getting good control then maybe you haven't configured it correctly? It could be poorly tuned, or have too long of a duty cycle, etc..

FWIW, in process industries at least, you wouldn't typically have an "electronic regulator" - your controller becomes the regulator, and you wouldn't have it controlling for pressure - you'd have it controlling for temperature. It would naturally adjust for input pressure changes as the tank empties.

It's indeed a PID but I'm using the alarm relais outputs, only. Of course the PID is able to do the PWM but how would you use it to control mechanical valves?

Does there exist electronical controllable (needle) valves which can use the PWM output directly? I think I will need a component between the PWM out and a motor valve and I'm planning to use an arduino therefor as I've got some lying around and doing nothing :-)

[Andy98]

Sorry it's taken so long to reply - I'm not sure what happened, I actually typed up a full reply quite a while ago and just noticed that it went missing. Oh well.

 

On 9/10/2016 at 5:28 AM, theimi said:

It's indeed a PID but I'm using the alarm relais outputs, only. Of course the PID is able to do the PWM but how would you use it to control mechanical valves?

Ok, just to get some terms out of the way, there are On/Off valves (aka shutoff valves, or digital valves) which are meant to be fully open or fully closed.

Then there are modulating valves (also known as control valves) which can be adjusted to between 0 and 100% open.

I think you're asking, how can you control a modulating valve from an electronic output. In industry, the answer typically is you use air pressure to actuate the valve, and you have another device that controls the air pressure in response to your output signal (typically a 4-20ma signal). Here is the wikipedia page on Control Valves.

It's less common, but motorized valves do exist. They are usually for special applications (I've only ever seen them for very large valves). Note that when you do motor controlled valves, you also generally have a position sensor as well.

None of the above falls into hobby-level pricing. With some quick googling I did see that some people go about building stepper motor controlled valves. I'm not sure I'd be super comfortable with that, myself, for dealing with propane.

For PWM, you'd have the actual pulse output going into the solenoid (or perhaps you'd put a relay between the two). So the solenoid might open for 30 seconds, then close for 30 seconds, then open for 30 seconds etc... depending on how you configure the duty cycle. In your application that would mean you'd have 30 seconds full heat, then 30 seconds idle heat, etc.. representing a 50% control output. 

Note that you need to control the blower as well as the fuel - so, using the above example again, off the top of my head I'd wire the PWM into a relay and have one contact from the relay wired to the blower to kick it into full speed, with another contact going to the gas solenoid. Again, I've never automated a forge so there might be better approaches. I haven't even researched the topic.

If you want to use the arduino, I'd be tempted to just ditch the PID controller and find an open-source PID arduino project. I'm sure there are dozens - I know that home brewers do that sort of thing a lot to control fermenting tanks. I would NOT write a PID controller from scratch - the basic idea is straightfoward, but there are bells and whistles that would be annoying to replicate.

EDIT:

Interesting, found some options on Ali Baba. This one includes open and close position switches, so you could home it against the close to figure out the position. This one just has an "open" command input and a "close" command input, so you'd need your ardunio to home the device and track what % open it is.

https://www.aliexpress.com/item/5-wires-CR05-DC12V-Electric-motor-valve-1-4-Brass-2-Way-Motorized-valve-DN8-Electric/32432458325.html?spm=2114.10010108.1000013.3.M3NzwY&scm=1007.13339.33317.0&pvid=ef2793c6-7f93-44cc-914d-1dc74d7be792&tpp=1

This one actually takes an analog 4-20ma input - so it must be doing the position tracking itself:

https://www.aliexpress.com/item/4-20mA-or-0-10V-AC-DC24V-control-1-1-2-proprotion-Modulating-valve-for-flow/990932550.html?spm=2114.10010108.1000010.2.IvmCiH&scm=1007.13438.37934.0&pvid=3eff91a9-1121-4d3b-8c78-d806d6a411b2&tpp=1

...that last one there says it's proportional, but the flow graph isn't linear. Basically if you open the valve twice as much, you want to be getting twice as much propane otherwise your PID will get confused (it wants a linear system).

Anyway, these would all be a shot in the dark. I did not evaluate them at all for this application other than to note that they are control valves. They are probably totally inappropriate for fuel gas or forge use - I'm just pasting the links to give a sense of what is out there.

 

 

[Latticino]

Don't have time right now for a full response, but here are a few notes:

1. electronic control valves for piping certainly exist.  Pneumatic to electric conversion is not essential, and even a bit out of date, though reaction times may be faster.  Modulating control valves rated for gas service may be expensive though.  I am more familiar with those for hydronics.  Usually they use a 4-20 MA signal or a 0-10 VDC signal for control.
2. Control authority is necessary to approximate the linear response you are looking for in the action to valve feedback loop.  This linear response is not critical (see below), but it does make things easier.  Typically this is done by reducing the valve diameter one pipe size relative to the supply line (i.e. if you are supplying at 3/4" your control valve is 1/2").
3. For gas it may be easier to have an "idle" circuit with a low level bleed serving your burner (manual valve and needle valve) and a "power" circuit with a high level bleed.  Both go into your single burner and the controller modulates between high and low fire rates based on feedback.
4. Actually I would recommend use of a proportional "zero pressure" regulator that keeps the gas and air mixture in the same proportion regardless of combined flow rate (within reasonable ranges).  Then you modulate the air to low and high flow rates and the gas follows suit.  Modulating air valves are much easier to acquire (and safer to use).
5. If you are only using a linear response feedback loop from a PID controller you have it configured incorrectly.  Do some reading on Proportional/Integral/Derivative  feedback control.  Sounds to me like you have your constants set incorrectly and are experiencing overshoot.  Setting constants up correctly is something of an art, but newer controllers have autotune capabilities that may make this easier.  Unless you are trying to reinvent the wheel programming your own controller sounds like a waste of time to me.  Programmable PID controllers for temperature control with ramp and soak functions are pretty available and cheap these days.
6. Thermal mass in the object to be controlled will make the response slower and more stable.  Generally speaking you need to have the heat source well sized for the volume and mass being heated to permit accurate and converging control strategies. 

[Andy98]

16 hours ago, Latticino said:

If you are only using a linear response feedback loop from a PID controller you have it configured incorrectly.  Do some reading on Proportional/Integral/Derivative  feedback control.  Sounds to me like you have your constants set incorrectly and are experiencing overshoot

FWIW he said above that he's actually got the solenoid valve wired into the alarm output on the PID, so it's basically just working as a temperature switch.

 

16 hours ago, Latticino said:

• For gas it may be easier to have an "idle" circuit with a low level bleed serving your burner (manual valve and needle valve) and a "power" circuit with a high level bleed.  Both go into your single burner and the controller modulates between high and low fire rates based on feedback.
• Actually I would recommend use of a proportional "zero pressure" regulator that keeps the gas and air mixture in the same proportion regardless of combined flow rate (within reasonable ranges).  Then you modulate the air to low and high flow rates and the gas follows suit.  Modulating air valves are much easier to acquire (and safer to use).

These seem like really, really good ideas.

Poking around for proportional regulators lead me to the "home-roasters" community (wow, there are a lot of hobbies out there..) and apparently those guys build PID controlled, gas-fueled, coffee roasters. Here is one guy's experiment: https://www.youtube.com/watch?v=X75Rx_mezHs

...he found an inexpensive proportional solenoid valve (not a proportional regulator) which was interesting (to me).