Tank heating

[th33ngineer]

Hi everyone, I am new to the site and I have a small design problem.
I work on a ship with a swimming pool that is drained to a retention tank when the ship is cruising at sea. I have been asked to install a heating element into the retention tank. Sounds simple enough. The problem is that the only 440V distribution panel that is close is for the laundry and has a lot of equipment on it already. I am limited by the size of cable that supplies this board from the main switchboard. The load on this board will vary depending on what equipment is running. Running a new larger cable is an expensive and time consuming option.
My thoughts were to install a thyristor power controller that could vary the power supplied to the heating element depending on the total load being taken by the distribution panel. This way when the load increased due to the running of other equipment the thyristor controller could reduce the load on the heater and thus protect the incoming cable and prevent the main breaker tripping on the main switchboard.
Most installations i have come across have the thyristor controller operating on the signal from a temperature sensor only. I have not found an installation that could not only monitor the temperature and control the heater but also monitor the current being drawn by the distribution board.

Any ideas?
Thanks in advance.

 

[Mike odom]

you could use a current transformer around the main wire to monitor the current draw, and if your controller was set to a max current (including the swimming pool), then it would automatically back off the heater when other loads were being drawn.

I did a board last year that monitored the current to a fan and turned on a relay if the current dropped below a certain point using a comparator, simple digital switch circuit (was actually 8 on a board). Designing an analog controller is a little more work, but very doable.

I have also designed and built temperature controllers for a waffle cooker in analog, but this circuit just turns the irons on or off and uses the large heated mass as the moderator.

Controlling a thyristor in analog is a little more problematic. Although you can control the average or rms power quite easily using pulse width modulation techniques, controlling actual power is a little more difficult. For example, the company I worked for built a 70 amp deep discharge battery charger system. Basically, we would hit it with 250 amps at a duty cycle of 28%, for an overall average of 70 amps. We hit it with 250 amps when we turned on, because that is what the source would supply, and regulated the average charging power by adjusting the duty cycle. Of course, we used a micro controller to monitor the charge and adjust the duty cycle. This would be a digital controller, not an analog controller.

What kind of power is feeding the mains?

here is an example of a current transformer...

View attachment CT1040.pdf

https://www.digikey.com/catalog/en/..._sub_type=Part Detail Page&WT.z_ref_page_id=0

 

[crutschow]

I suppose there's no chance you could route some of the engine cooling water to a heat exchanger in the tank?

 

[Diver300]

you could use a current transformer around the main wire to monitor the current draw, and if your controller was set to a max current (including the swimming pool), then it would automatically back off the heater when other loads were being drawn.

With that approach you are in danger of getting oscillation. If the current draw is less than the maximum, and the tank heater takes it over the maximum, then as soon as the heater turns on, the current goes over the maximum, and the heater gets turned off.

I suggest that you measure the rest of the loads, and allow the heater to turn on when the current small enough to allow the heater to run. There is no risk of oscillation like that.

For instance, if the supply is 50 A, and the heater takes 10 A, then allow the heater if the current taken by everything else is less than 40 A. When the heater turns on, the rest of the loads must be less than 40 A so there is enough spare capacity for the heater.

I assume that you have some way of monitoring the current anyhow, for commissioning? You need a thermostat as well, but I would hope you know that.

Routing the water to a heat exchanger to the main engine cooling water would save fuel. It still needs a thermostat. I've seen a system like that with a pool that got way too hot because the thermostat was installed incorrectly.

 

[picbits]

Routing the water to a heat exchanger to the main engine cooling water would save fuel. It still needs a thermostat. I've seen a system like that with a pool that got way too hot because the thermostat was installed incorrectly.

Personally I'd use a mechanical thermostat set at say around 50 degrees (for safety) in series with an electrical operated valve which you could control electronically for a finer control.

If either failed the other would act as a failsafe.

If you have excess wasted heat from the engines, it would be a shame to waste it

 

[th33ngineer]

The engine jackets are sea water cooled. In any case this would involve running pipework through fuel tanks. Trust me that was the first and easiest idea I looked at. It just isnt viable. Not only that, a failure of the heat exchanger could result in loss of the engine and contamination of the pool system.

Diver300 - To prevent oscillation, that is why I am looking at thyristor control. The power that the heating element would be allowed to draw would be controlled by the thyristor power controller, this would be proportional to the total current draw on the board.

Mike - CT's are exactly what I was looking at. I already use CT's onboard for monitoring earth leakage on 3 phase plant. I am not investigating the use of CT's and max current controllers like you mention. Integrating this to a temperature controller is where it is becoming tricky. I am probably trying to make it more complex than it needs to be. A PID temperature controller is probably the way forward.

 

[Diver300]

You don't need a PID controller. You just want an on/off controller. Will a big water tank it will heat up so slowly and so uniformly that there really won't be any overshoot.

If you only have on/off control, it makes switching the heater much simpler. Thyristor control is an extra complication that you don't need. If you are really short of power and you absolutely need to maximise the power that you can get to the water tank, you might need it. However, measuring the current, and working out how to control the thyristors to get the right RMS current isn't easy.

Unless the heater is going to use most of the available power, I would guess that there will be spare capacity most of the time.

 

[th33ngineer]

The problem is that the on/off control will constantly be bringing on and taking off quite a large load for this board. The load on the board will constantly vary as other consumers switch on and off. Using a CT to control the heater for max current would work along with a on/off control from a thermostat..... i think

 

[crutschow]

How much current will the heater draw at maximum? Is this 50/60Hz?

Edit: Instead of using thyristors, which can cause EMI issues, how about using several heaters and connecting them in serial or parallel to control the power?

 

[th33ngineer]

I cant answer that yet as I am still waiting on information on the size of the heaters required. I am guessing that there is prob be two 18KW heaters. This is 60Hz, 440v.

Connecting the heaters like that would mean that the control would be non variable. The power control needs to be proportional to the total load current on the distribution board.

 

[atferrari]

Out of curiosity, may I know what a "retention tank" is? An equivalent of a "settling tank" for bunkers?

Where the water goes to, after leaving the tank?

 

[Lourens]

The attached control logic can be used for your pool water. The control can be implemented with a PLC or programmable controller.

 

[Mike odom]

With that approach you are in danger of getting oscillation. If the current draw is less than the maximum, and the tank heater takes it over the maximum, then as soon as the heater turns on, the current goes over the maximum, and the heater gets turned off.

wrong again...
you monitor total power, including the power you are supplying, for a maximum, and as other loads turn off, you turn your power out up, and as other loads turn on, you turn yours down. This is done all the time in electronics. This is also why breakers and fuses have time delays, they don't blow the microsecond their current rating is overshot.... the trick is to supply average or rms power overall, and to keep those overshoots to a one cycle minimum. And the output is already oscillating, at 120Hz... ) (60Hz has 120 zero crossings per second).

Also, breakers aren't specified by the load. They are specified to protect the circuit, ie, the wiring. If the load shorts, the breakers will trip BEFORE the wires melt, or reach a temperature that would start a fire. There may be 60-100% cushion (should be at least 50% if designed right... you don't put a 50A breaker on a 50A load, you use wiring that would deliver 100A and a breaker set at 75A).

 

[Mike odom]

I suppose there's no chance you could route some of the engine cooling water to a heat exchanger in the tank?

This is actually a very good idea... so you can't use engine coolant, how about the waste heat from the engines going through the stack?

 

[Mike odom]

Out of curiosity, may I know what a "retention tank" is? An equivalent of a "settling tank" for bunkers?

Where the water goes to, after leaving the tank?

I'll take a stab at this...
With a ground pool, you fill it and drain it to the city sewer... on a ship, fresh water is at a premium, so to drain the pool you don't want to through it overboard so you pump it to a retention tank. When you are ready to refill the pool, you pump it back to the pool. It seems they want to keep it warm so they don't have to wait for the water to warm up once they refill the pool.

Too bad the retention tank isn't in the engine room. Ours was a brisk 120°F on the submarine... we did pump about 15000 gallons of fresh water overboard a day though... our fresh water generators were finicky so we ran them at full and discharged the overage. Better to have too much than not enough...

 

[Diver300]

With that approach you are in danger of getting oscillation. If the current draw is less than the maximum, and the tank heater takes it over the maximum, then as soon as the heater turns on, the current goes over the maximum, and the heater gets turned off.

wrong again...
you monitor total power, including the power you are supplying, for a maximum, and as other loads turn off, you turn your power out up, and as other loads turn on, you turn yours down. This is done all the time in electronics.

I had made the assumption that the control would be on/off control. If you only have on/off control, and the overall current is measured, it will oscillate.

Proportional control will work fine, and will keep the power near the maximum rating of the circuit. However, it is quite complicated, and it is quite expensive to design proportional control for a couple of 18kW heaters.

On/off control is far simpler. Whether anything more complicated is needed depends on details of the installation, mainly on how much of the available power will be needed to keep the tank warm.

My guess is that the galley load will be very variable, and that there will be long periods of time when it is far from its maximum. If the heater is also considerably less than the maximum power, then on/off control will be fine.

 

[Diver300]

I cant answer that yet as I am still waiting on information on the size of the heaters required. I am guessing that there is prob be two 18KW heaters. This is 60Hz, 440v.

Connecting the heaters like that would mean that the control would be non variable. The power control needs to be proportional to the total load current on the distribution board.

I still don't see why you will need proportional control. Most temperature control of water tanks is on/off. The variation in temperature is generally only slightly more than the hysteresis of the temperature switch.

What is the rating of the circuit that feeds the laundry area? Do you know what the usage profile is? How big is the water tank?

Just as a thought, make sure the heater is off when tank isn't full enough to cover the heaters and the temperature sensors.

 

[crutschow]

I still don't see why you will need proportional control. Most temperature control of water tanks is on/off. The variation in temperature is generally only slightly more than the hysteresis of the temperature switch.

...............

He needs proportional control because his concern is about overloading the breakers. On/off control will (momentarily) apply the full load to the breaker, which would rapidly open the breaker if the panel is overloaded.

 

[Diver300]

He needs proportional control because his concern is about overloading the breakers. On/off control will (momentarily) apply the full load to the breaker, which would rapidly open the breaker if the panel is overloaded.

I was suggesting an on/off control with an inhibit when there isn't enough spare power. How well that would work would depend on how often the heater would have to be inhibited.

In house heating, it is quite common to shut off the heating when hot water is needed. Most house owners don't notice.

 

[th33ngineer]

This is actually a very good idea... so you can't use engine coolant, how about the waste heat from the engines going through the stack?

This would involve a lot of structural re-design. Also as we have four engines a design like this would become very complex if we were to be able to use any engine for heating. This is not a viable option.