Cooling down a steam boiler

Harvey Ramer

What is the best way to get rid of the remaining steam in a process application?

Low pressure steam heats water through a giant hx.
Water runs through enormous plate exchanger to heat milk/cream to a certain temp.

What is the best way to end the process? Does it require a btu dump system?

Jamie Hall

What is the actual need here, @Harvey Ramer ? Is it to actually cool the boiler, or is it (as I have a sneaking suspicion it might be) to rapidly transition the milk or cream from pasteurization down to bottling temperature, avoiding as much as possible a slow drift down through potentially dangerous warmer temperatures?

If it's the latter, I've simply switched the water flowing through the enormous plate exchanger from the hot source (that steam/water heat exchanger) to a cold or at least cool source (a big enough cool source can be hard to find, unless you can use the "waste" heat profitably some way -- back in the day I used a nearby stream and just dumped the water back. Very wasteful!). Then turn the boiler off or just close the steam supply valve to the steam/water heat exchanger.

Harvey Ramer

I should have been more specific.
When the process is shut down or ends and the hot boiler is turned off but is still steaming, what to do with the steam till the boiler cools down enough to quit producing steam?

Harvey Ramer

The product side of the process hx can easily be controlled by varying the speed of pump between the 2 hx's.

Jamie Hall

You're going to have to dump the heat in the steam somewhere... if you don't want to waste it, keep the steam/hot water HX on and use the resulting hot water as a preheat for domestic hot water, or space heating, or...

Harvey Ramer

@Jamie Hall Got it. Now on to the next thing.

Again, the current setup = a 1.7 mbtu/h boiler with a 3:1 mod gas burner. Steam set at 5 psi. Steam heats water through a giant hx. Closed loop water pumps through and heats massive flat plate hx.

The product moves through flat plate @ fixed speed. The hot water loop operates at fixed speed. They have a modulating valve on the steam supply to the giant hx. It modulates based on the target outlet water temp in the closed loop.

I watched it for a while and noticed the modulating valve never stops modulating. Sometimes it even closes completely. Neither does the boiler water settle down. It surges up to the top of the sight glass.

I don't like the modulating valve and propose to remove it. How then would be the best way to control the closed loop water temp?

ethicalpaul

Sorry for the amateur opinion, but I love puzzles!

Can you circulate the closed water loop in such a way that you use a thermo mixing valve so that the closed loop flows through the hx (or partially does) when it needs heat and avoids the hx when it needs no heat?

Then the modulation happens with the mixing valve and the steam boiler will just cycle on pressure when it isn't needed.

This would be like putting a hot water loop on a steam boiler: https://heatinghelp.com/systems-help-center/how-to-run-a-hot-water-zone-off-a-steam-boiler/

Harvey Ramer

But the flow rate in the steam hx has to be maintained to keep the water from boiling?

Perhaps the burner should be modulated based on the water outlet temp of the hx instead of pressure? But would that be responsive enough?

ethicalpaul

Ahh I see what you mean, thank you Harvey.

Although, if you kept the water in the steam-to-water HX at something a little north of 5psi, you'd prevent the water from boiling, right?

pecmsg

Sounds like your control valve is under / overshooting. What’s the COV setting on the valve?

Harvey Ramer

"COV" ?

pecmsg

Change of value.
Hue quickly it responds.

Harvey Ramer

pecmsg said:

Change of value.

Hue quickly it responds.

Right.
I don't know. It is tied into the purpose built process control system.

But can a modulating steam valve even work decent in this situation? Considering the pressure temperature characteristics of steam and how the hx will start inhaling air every time the valve closes completely.

Jamie Hall

You have too big a time delay between the change in temperature in the water circulating through the process and the effect of the control -- and with that boiler sitting there happily producing steam and having a very slow response it's not surprising that the system oscillates -- the control valve modulates all the time. Again, I'm sort of speculating here a little, but I would imagine that the desired result is some approximately constant temperature of the product at the output of the heat exchanger? This is would be controlled in turn, by the flow rate and temperature of the water on the water side of the plate exchanger. I think the first thing I'd try to pencil out would be, as @ethicalpaul suggested, a mixing valve at the water input to the plate exchanger, controlled by a temperature sensor on the product output. The two inputs to the valve would be the steam/water exchanger and a simple bypass line (perhaps with a throttling valve in it, to keep the head loss the same no matter where the missing valve was in its travel -- that would prevent flow rate variations). And I'd probably shoot for having that mixing valve at steady state somewhere in the middle third or so of its modulating range. Then I'd keep the modulating valve on the steam/water heat exchanger, and control it based on the outlet temperature of that heat exchanger. That control signal would also control the modulation of the boiler, although the modulation could be on boiler pressure too, though that's more time delay than I'd like to see.

To cope with the heat delay when the process shut down, you could keep the steam/water heat exchanger, but divert all of its output away from the plate exchanger to something vaguely useful while the boiler came off the boil...

Time delay is the enemy of close control -- anytime there is a time delay in a control loop between action and result, you run the risk -- almost inevitable -- of oscillations like you are seeing.

pecmsg

Has it ever worked properly?

Perhaps lowering the steam pressure. It won’t heat the product as fast.

Harvey Ramer

pecmsg said:

Has it ever worked properly?



Perhaps lowering the steam pressure. It won’t heat the product as fast.

No, it hasn't. It's a new yogurt plant and they haven't scaled up to full production yet.

The boiler falls flat on it's face during startup when they are bringing in 35° milk and then cycles a lot during pasturization.

Harvey Ramer

Jamie Hall said:

You have too big a time delay between the change in temperature in the water circulating through the process and the effect of the control -- and with that boiler sitting there happily producing steam and having a very slow response it's not surprising that the system oscillates -- the control valve modulates all the time. Again, I'm sort of speculating here a little, but I would imagine that the desired result is some approximately constant temperature of the product at the output of the heat exchanger? This is would be controlled in turn, by the flow rate and temperature of the water on the water side of the plate exchanger. I think the first thing I'd try to pencil out would be, as @ethicalpaul suggested, a mixing valve at the water input to the plate exchanger, controlled by a temperature sensor on the product output. The two inputs to the valve would be the steam/water exchanger and a simple bypass line (perhaps with a throttling valve in it, to keep the head loss the same no matter where the missing valve was in its travel -- that would prevent flow rate variations). And I'd probably shoot for having that mixing valve at steady state somewhere in the middle third or so of its modulating range. Then I'd keep the modulating valve on the steam/water heat exchanger, and control it based on the outlet temperature of that heat exchanger. That control signal would also control the modulation of the boiler, although the modulation could be on boiler pressure too, though that's more time delay than I'd like to see.

To cope with the heat delay when the process shut down, you could keep the steam/water heat exchanger, but divert all of its output away from the plate exchanger to something vaguely useful while the boiler came off the boil...

Time delay is the enemy of close control -- anytime there is a time delay in a control loop between action and result, you run the risk -- almost inevitable -- of oscillations like you are seeing.

At this point there is nothing else being done by the steam boiler.
Might it not be better just to pump boiler water direct to the process hx and put a drive on the pump to control the speed. Basically use it as a hot water boiler.

Mike_Sheppard

Having a control valve for the steam heat exchanger is fine. But it needs to be sized and controlled properly. I would suggest having a 1/3 2/3 valve setup where one valve is sized to provide a third of the load and the second provides two thirds. The 1/3 valve opens first and the 2/3 valves opens if needed. If you have one big steam valve and it’s hunting up and down it’s going to wreak havoc on your water level.

If you have enough pressure on the water that the steam is heating, it won’t turn to steam. We have applications here in tall buildings where they use water to heat the building but use steam boilers to heat the water, this way they don’t have to pay extra for high pressure rated boilers. They run 10 psi steam which is constantly heating the water up to around 240 degrees. But they have around 80 psi water pressure on the system which means the water is not going to turn to steam.

Assuming your steam boiler is low pressure with a maximum of 15 psi, all you would need to do is keep your water pressure above 15 psi and it won’t turn to steam. Then you could use a 3-way valve on the water side to maintain the temperature of the water or product.

That being said, if you’re boiler and everything else is extremely oversized you’re only fix for that will be adding more load or using a large buffer tank.

pecmsg

That’s your issue. Design calls for X but never figures 1/2 or less of that. @Mike_Sheppard has a good idea. Repipe the control with 2 valves. High and low.

Jamie Hall

Harvey Ramer said:

At this point there is nothing else being done by the steam boiler.
Might it not be better just to pump boiler water direct to the process hx and put a drive on the pump to control the speed. Basically use it as a hot water boiler.

Yes. If the flow rate of the milk/cream varies much, or its input temperature does, you might want to add some predictive control on the pump based on mix input temperature and flow, and trim the flow based on output temperature.

EBEBRATT-Ed

@Harvey Ramer

Yes, @Mike_Sheppard is right

going 1/3 2/3 on the valves will help.


Why did they use a steam boiler instead of a hw boiler?

Temperature is my guess

Your existing valve is hunting and the wide swings in steam flow is causing an unstable water line.

The swing in pressure/temperature can also be caused by the valve trying to move to fast.

to fast or to slow on the valve will cause the same thing, it's trying to catch the load.

The other problem and this is a big problem is the load is unstable (because there isn't enough volume of water) and the valve can't catch up.


I have had this problem with chillers you need 6 gallons of water volume in the system/ton for a process system and 3gallons/ton for comfort cooling. The reason is process loads change fast and comfort coolin is much more stable

I think this is the same problem your having

I would but a big buffer tank between the two HXs

This will give you stability. I think that's the sure fix