Chilled Water Design Engineer problem

Hi Guys,


I am after some engineering information regarding a chilled water project,

Basically we have a food processing company that wants to use chilled water for process,

the water being supplied to process is not returning back into the chilled water tank, it is going to waste, the flow required to 80L/min or 1.4 L/sec,

the mains makeup water is at 24 Deg C and was going to use a single chilled water tank, I have attached a schematic of piping arrangment,

I am after some calculation/formula to work out what the heatload will be on the chiller,

Eg: If I have a tank that is 3000L capacity, and the incomming water temperature is at 24 Deg C (mains water makeup) and I want to start the chiller 3hrs before start up then the load is 23 kW however when productions starts how do I work out the tank temperature or load when 24 Deg C water is mixing with say 7 Deg C water?

If i work it out on water entering at 24 Deg C into the chiller and comming out at 7 Deg C (ideal supply temperature) then my kW is 112 kW based on a flow of 1.4 L/sec.

We want to get stratification inside the tank,

Also is there a general formula to calculate chilled water tank temperature?

I want to make sure it is not oversized for the job.

Any help much appreciated

PS : I tried to upload drawing but cannot exceed 100kb and file is 350 kb, let me know and I can email

Quote:

---

Originally Posted by freezerman11

Hi Guys,
I am after some engineering information regarding a chilled water project,

---

Hi, Is this homework? I’ll try to help without doing it for you.

Quote:

---

I am after some calculation/formula to work out what the heatload will be on the chiller

---

You worked out your example soy you have the formula, you must estimate losses though.

Quote:

---

We want to get stratification inside the tank,

---

This is not possible with water, you can get it with air because there is little density difference, but with water you will get natural convection inside the tank, besides with that water flow it is impossible not to create currents in such a small tank.

Quote:

---

Also is there a general formula to calculate chilled water tank temperature?

---

No. It will be close to a weighted average temperature. But you can use your imagination (maybe this is too much help): You can split the tank in two sides, hot and cold with an overflow in between, thus you do not need to worry about the mixing of flows, unless you overflow one side.

Quote:

---

I want to make sure it is not oversized for the job.

---

One of the main problems in the real world is to match the capacity of the system with the required performance. This is so because information in the real world is uncertain.

Are you sure your process will never input more energy?

Are you sure all your calculations are correct? If your answer is yes it means you integrated the heat capacity of water over all the temperatures it will take between initial and final states, or did you use an approximation?

Are you sure you calculated your losses exactly as they will occur?

If you are sure then you do not need to oversize and If I were your teacher, would flunk you!

Good luck!

You are based in UK?
And propose to waste fresh water at a rate of 80 l/s ???

No wonder we got a hosepipe ban...
:rolleyes:

.

Potable process water requires double wall separation between refrigeration circuit and process water. A standard water chiller, filling a tank won't give you that.
A method to acheive result would be use closed loop through chiller and 3000 tank and maintain loop temp., and then use a plate heatchanger, with secondary side feed from tap water, the outlet to the process.

( the specific heat of water is 4.186 Kj/Kg )

As the Viking asks quite rightly, why is your process water going to waste?

Hi all,

thanks for the feed back,

Viking, the water flow rate is 80L/min not 80L/sec = 1.4L/sec...

The water cannot return back into the tank as it is contaminated after going through process...

when running production the mass water flow going to production is 1.4 L/sec say therefore the makeup water needs to supply water to tank greater than 1.4L/sec , however if the tank was to be 80% full already with 7 Deg C water, would the heatload calculation be based on 24 deg c entering or 7 Deg C entering the chiller, as there would be some sort of thermocline in the tank...

Cheers

Can you please explain your process and application, as the solution lies within the application.

Hi Mad Fridgie,

The process is for fish processing, currently they use ambient water at say wetbulb average 24 Deg C,
the water cannot return back into chilled water tank as it is extemely contaninated...

The water is supplied to filleting bays, this water at present is 24 Deg C, then goes to waste, however we are looking to chill this water at 7 Deg C, this eliminates mirco organisms / bacteria etc generating on the fish.

My question is that the mass flow required to the filleting bays is 1.4 L/s, however if we were to have a chilled water tank that is chilled to 7 Deg C before production was to start, and the water mains makeup float is suppying water at the same rate/or greater than what is being supplied to production, what would I calculate my heatload on? would it be 24 - 7 Deg C? giving 110 kW or would it be less because the tank is already fully of 7 deg C water and mixing of 24 Deg C with 7 Deg C?

Cheers!

Quote:

---

Originally Posted by freezerman11

Hi Mad Fridgie,

The process is for fish processing, currently they use ambient water at say wetbulb average 24 Deg C,
the water cannot return back into chilled water tank as it is extemely contaninated...

The water is supplied to filleting bays, this water at present is 24 Deg C, then goes to waste, however we are looking to chill this water at 7 Deg C, this eliminates mirco organisms / bacteria etc generating on the fish.

My question is that the mass flow required to the filleting bays is 1.4 L/s, however if we were to have a chilled water tank that is chilled to 7 Deg C before production was to start, and the water mains makeup float is suppying water at the same rate/or greater than what is being supplied to production, what would I calculate my heatload on? would it be 24 - 7 Deg C? giving 110 kW or would it be less because the tank is already fully of 7 deg C water and mixing of 24 Deg C with 7 Deg C?

Cheers!

---

Is the process constant (flow), per sec, min, hr (how long per 24 hrs) and controls on flow apart on and off

hello freezerman
must know that in order to reduce by 15 ° C to 14 ° C n ° 1 liter of water in 1 hour serves a frig / h.
then according to the temperature difference and the quantity is simple to make the calculation.
remember to also add to the cooling load losses due to thermal insulation, pumps, the actual load of the user.

You need to put in a tank big enough to deal with daily usage, end of. This tank would be chilled over night and opens you up to using cheap rate electricity.

al

Al,

If we were to do this the tank volume with be of great capacity, the larger the tank volume means smaller capacity the condensing system is and vs versa, this would be a tank of approx 40,000 litres

This is more a mixing water temperature calculation of two temperatures being 24 deg c and 7 deg c

Cheers

A site i look after has a tank of over 100000 liters, sufficient for a week of running,the tank itself acts as a huge thermal mass with hot/cold side and overflow, fair enough if you don't have the physical space. So what is your own background? design or service?

can you not find a way to decontaminate the water to allow for reuse? Here we are worrying about global warming and yet we have this kind of wastage, foolish.

Fish plants i know of here just use sea water, no chilling, filtered and returned to sea.

al

If you have a single drain point, then you can use add a heat exchanger between the 2 flows, (a long tube in tube, with CIP points for cleaning)
The tank set up all depends how the evap has been designed to meet the duty, is it a single pass design, 24 to 7C or is higher flows with smaller TDs?

If this is fresh water.

If you did not have a tank then your chiller would need to be (Flow rate in litres@sec x delta temp x 4.2) 100kW. You now say that you have a 3000l tank. So for an 8 hour day you will need 40320lites less the 3000 that you have stored in your tank which = 37320. So for every litre of water you take from the tank you will need 12.44 litres from the mains. So therefore every second you will need 0.105l from tank and 1.295l from mains therefore the chiller will need to be 92.5kW. Not too much of a kW saving and you will not save energy overall from a smaller chiller only thing saved might be initial overall cost. Then one must take into account the cost of the installation needed to make the system work. This is a rough idea of how to take on the calculations. If you need further detailed design then let me know I have done a few similar type of projects in the past and may be able to help.

I think you will find that the chiller is somewhat standard, where the chiller flow and return temps are some what closer together, where the chiller duty is produced at at much higher flow rate, the use flow rate. The tank is being used as a mixing chamber and short term thermal buffer. (anti cycle protection)
I think the OP concern is that when introducing make up water it will influence the temp in the tank. That being the case he want to know where to place piping into the tank.
For this we need to know the chiller design flow rate and the deviance allowance of the supply water temp to the process.

Hi MF.
suggest a tall narrow buffer tank and introduce water make up horizontally at the top, creating a whirlpool effect, the temperature gradient top to bottom will be slower, also suggest a weir up the middle top to bottom and standard chiller drawing off warm side at high level and reintroduced back into colder side at low level. The greater the volume of tank the better and will reduce the output temp slippage over a days production.
I am concerned with using a standard chiller and possible contamination of the potable water with single wall separation between refrigerant and process water. i have heard of a daily production load of suniso oil flavoured bread having to be dumped due to a plate HX failure. Won't mention where that hails from.
magoo.

Tall and skinny all good with me, I would not go split buffer on this application, not being a true recirc system.
I was think of introducing the make at the top or even in the chiller inlet (chiller draw off at the top)
Stat about 1/3 the way down, chiller return through a sparge about 1/3 the way up.
Process draw off at the bottom. Gives some thermal lag, and aids is efficiency, by keeping SST as high as possible.
Stat say set a 6 off and 8 on. Should give a slightly lower them than required. But all comes down to the the chiller split.

Hi MF
you are probably correct, my idea of the baffle was to get further separation and maintain max duty from low capacity chiller and highest SST.
No mention of potable water separation, there is nothing like oily fish fillets on the BBQ. Contamination of product could kill a company in one production shift.
magoo