Help Designing Shell and Tube HX

This might help...

https://link.springer.com/content/pdf/10.1007/s11434-012-5558-4.pdf

Thank you!

Should show us what you already have... remember to transfer heat... you have to comdense the steam to a liquid by using a impingement baffle.

And how do you figure out the ΔP on the shell side without empirical data from actual tests?

hey, this is the spreadsheet with what I have so far.

Those are unreadable here.

Whoever gave you the data included several superflous items and omitted many necessary items.

I'm not sure if that's where we're supposed to 'Think'. But I've not been able to get a low pressure drop for my shell-side regardless of what I try. I was looking at the cooling water exit.... It's 140 degrees C! I asked if there was a mistake there but was told that there wasn't any. I really am lost at this point.

The water has to take the enthalpy of the steam, 22,000 lb/hr, to change state, more BTUH than changing temp of water...

That's the part you need to look up in the steam tables, some 919 more BTU per pound of water/steam than required to heat the water, or some 20MMBTUh of heat has to move. Maybe 200 gallons per minute of water, some 200F rise?

Assuming you don't let the water flash to steam....

Shell side pressure drop is the steam side, perhaps you are not converting the volume of steam in to volume of water out, which will significantly reduce the velocity at the exit, bringing the calculated pressure drop in the Hx casing more into line...

There is some weird stuff going on here that doesn't add up but lets try and see what we can do.

My first thought is what arrangement do you have for the exchanger. Surface condensers have a very specific layout.

The steam is in the shell and the cooling water in the tubes.

Steam enters from the top and there is ONE shell side pass with condensate leaving from the bottom. There is an impingement plate just larger then the nozzle set about a nozzle diameter below the nozzle which protects the top rows of tubes from the steam directly blasting on to them.

Otherwise there are NO baffles. If you have baffles then this will increase pressure drop.

You could reduce the pressure drop by making the tubes longer - that way fewer tubes would be needed so a smaller bundle and therefore lower pressure drop. From memory 6m is standard but you can have any length really (you'd pay more for it in real life but this is a design exercise)

In the real world cooling water temperature rise is limited to about 10C.

By allowing such a crazy temp rise your teacher has reduced LMTD increasing the surface area required for the heat exchange. So more tubes etc and so a bigger shell which has a larger pressure drop - so again longer tubes.

Also I get 145C as the condensate temperature. I would expect the steam to superheated exiting the turbine but the condensate should be saturated or a little sub cooled. This could have an impact on densities but most of the pressure drop is in the steam phase and the conditions for that seem plausible. Again just check you are using the right densities and viscosities.

Then of course there are units and conversions. Kern is a great book but I wish they did one in proper units! Its really easy to get a conversion the wrong way round or to miss out factors of 1000 or 1,000,000 etc. I would definitely double check all of those.

When you're talking layout, what are you talking about... process layout?, Tubular layout?

For surface condensers, These would be the baffle arrangements and I mentioned earlier. The baffles accomplish (2) things, to get the dispersion on the tubular surfaces, and as a impingement, to release the heat during the phase change.

Included in the link later on.

I take it tubular layout. If you look up TEMA requirements, this would have more.

For Tubular layouts for tubesheets, you have (4) options.

30° Triangular, 60° Triangular, Square, and rotated square.

The square layouts are required where it is necessary to get at the tube surface for mechanical cleaning. The triangular arrangement allows more tubes in a given space. The tube pitch is the shortest center-to-center distance between tubes. The tube spacing is given by the tube pitch/tube diameter ratio, which is normally 1.25 or 1.33. Since a square layout is used for cleaning purposes, a minimum gap of 6.35 mm (0.25 in) is allowed between tubes.

Here's is the link

Thank you for the link, yes it was the exchanger layout (tube bundle etc) that I was referring. My reasons for saying that this would have no baffles follow this logic

If you look down the Shell Types you will notice that Type X is specifically listed for this service and does not have any baffles, unlike the similar G, H type shells.

In most exchanger services baffles are a critical item for all the reasons you noted. However in this case due to the very low allowable pressure drop there is an exception and baffles are not used.

Having said all of this most of this requirement is based on the usual surface condenser at the end of the power drive that operates at some partial vacuum (depending on the cooling service available) for which very low pressure drops are required. In this case we have a low pressure 3.2barg, for which a higher pressure drop might be allwable in which case a baffled exchanger would give a smaller shell for the service.

If the exchanger is not operating under vacuum then the ejectors are not required to bleed out incondensibles this can just be done with a control valve.
I do wonder if the pressure drop requirement is overstated in this case.

Obtain a set of Steam Tables, for there is nonsense in the data as presented:

• For cooling water to exit at 140degC there has to be a number of bars of pressure on it and whatever is in the cooling water that doesn't evaporate will be left behind either on the heat exchange surface or at the point where the pressure reduces to ambient.
• If the discharge temperature of the condensate were to be 156degC then it will also have a considerable pressure on it. It too will flash to steam wherever that pressure becomes ambient.

This isn't a steam condenser heat exchanger as depicted at present: it is simply a device for releasing huge clouds of steam to atmosphere and wasting its energy content. One might well do away with the heat exchanger as a concept, and just vent it instead.

Its a process issue.... not seeing the P&ID, one would have a steam trap, unless it is a pass-through system.

secondly, its an excel sheet,... I had written a HX sizing program in Excel and Excel does not have high level programming that require iterations for solving, unless, VB programming is used in Excel.

The 140degree discharge temperature of the water doesn't make sense to me.

I worked with naval steam system condensers and typically the shell has an air ejector attached that keeps the condenser steam side at a significant vacuum. The steam enters, contacts the tubes and phase changes to water, which effectively removes that volume from the system. Any noncondensable gas resides in the shell and eventually acts as a block to steam flow. The air ejector pulls noncondensable gasses out of the shell and establishes the vacuum. The temperature of the steam inside the shell is determined by the vacuum level established by the air ejector, the temperature set by the condenser case pressure and read from a steam table.

I have seen circumstances where the vacuum in the condenser was such that the steam went to solid phase and blocked up the condenser with ice. The condenser had to be idled and warmed to defrost it.

I'd rerun your calculations with the exit area of the steam side being set as the exterior surface of all the tubes, as that is where the volume of the steam goes away as it phase changes to liquid water. Also plan on the exit conditions at the tube surfaces to be at the vacuum level established by the air ejector.

It's interesting that the condenser shell isn't designed to contain pressure, but rather to avoid imploding due to the interior vacuum.

The thing about naval condensers is that there is a virtually inexhaustible supply of cooling water available - the stuff in which the boat/ship is floating. It could be that the original poster doesn't have that as a resource.

Even so, most condensers operate at a vacuum and use an air ejector to strip noncondensable gasses. The temperature is set by the condenser vacuum and the effective exit for the steam is the tube exterior. The condenser vacuum does a lot to maintain the plant efficiency.

This was true on navy platforms as well as at a commercial plant I worked at.

The condenser pressure is determined by the temperature of the water within it. For at that temperature, when a vacuum is pulled on it, the water boils to limit the vacuum until there is no water left.

I'll meet you half way. Both are true. The temperature is limited by the pressure. The pressure is limited by the temperature. It's a 100% saturated steam environment and the temp and pressure work in lockstep. At 212 deg F the pressure will be 14.7psia and the surface area of the tubes can be considered the steam exhaust area of the condenser at that temperature and pressure. If the water entrance temperature is 32 deg C, then the cool end of the condenser tube near the water inlet will tend towards 0.69 psia and will be a localized cool spot and low pressure region. At the hot end of the condenser tube the temperature is 140 deg C and the pressure will tend towards 52 psia. In fact, the condenser measured ambient temp and pressure should be closer to the exit temp of 140 and pressure of 52 psia, but there will be a majority of steam condensed at the cool end of the tube and a predominant mass flow towards the cool end of the tube.

An accurate model would account for the localized temperature of the tube, the localized pressure and the local surface area for thermal exchange and mass flow rates for the phase change. An air ejector is still needed to prevent accumulation of noncondensable gasses.