17 replies to this topic

[Eprocess]

Dear experts,


I should propose the operating conditions of a steam generation facility, which is going to support a plant that needs just one steam level (saturated @ 6 barg). If you would have been asked to outline the operating conditions, what would that be? I understand that steam is generated in higher levels and then is regulated in desired levels, and I was wondering is that something that is done in every situation (because probably the boiler would be smaller) or is it rational to generate the steam you need at the desired operating condition?

My question does not concern issues such as pressure drop or heat loss, but I want to know what they would normally do in industrial plants for such a case.

Thank you

[sheiko]

Hi,
Having no superheat could lead to condensation. Usually, designers choose at least 10 to 20 degC superheat in the network to avoid condensation.

[Dacs]

It's more economical to have one single boiler that produces high pressure (HP) steam and provide the pressure letwdown (with temperature regulation aka attemperation) to medium (MP) or low (LP) steam, but this is for a plant that would require HP, MP and LP steam in the 1st place.

In your case however, it makes sense to design the steam generation unit at the expected steam condition dictated by the end user.

Hi,
Having no superheat could lead to condensation. Usually, designers choose at least 10 to 20 degC superheat in the network to avoid condensation.

Just to add on what he said, you also need to check the end user if it can tolerate a certain degree of superheat because doing this scheme will require you to have a desuperheater at the end user (since you require a saturated 6 barG steam), and even so, it may (depending on the type) have a certain amount of superheat (usually 5°C or so).

Normally however, you want to have your steam as superheated to be able to distribute it to your steam network.

As for the steam condition, you'd want to have some working margin for the pressure as your steam generation unit may fluctuate. I have some figures but I can't just outright recommend it as your mileage may vary (so to speak).

[breizh]

Hi,
Any indication of the process you want to serve ?

Batch reactors , distillation column ,....

Breizh

[Eprocess]

breizh:

We do have four plate and frame heat exchangers (main users, which regulate copper electrolyte temperature pretty tight), and a betch copper leaching reactor.

[Eprocess]

Dacs:

Thank you for the reply. I would appreciate if you tell me the margins that you would normally consider. I will not use those for my system, before studying my own case. But knowing them can help me a lot in evaluating my network.

[Dacs]

Sure with that. I had one project who uses 3.5 barG LP Steam. We designed the system to work on 3.1 (min) to 4.1 (max) barG.

By the way, looks like the end user (heat exchanger) cannot tolerate superheat. If it's possible to place the steam generation unit near the end user, you can do so without superheating the steam and just compensate by placing appropriate steam traps and insulating the piping, otherwise, I'd suggest a water-bath desuperheater to produce saturated steam if you opt to go for superheated steam generation.

[kkala]

In addition to the margins advised by Dacs, following data could give a picture, concerning two steam levels in a refinery. Steam supplied to network has some margins over steam received by consumers. However it is noted that mentioned levels of refinery steam does not come from ordinary steam boilers (generating steam of 42 kg/cm2 g), but from exchangers, reducing stations, probably back pressure steam turbines.
1. Steam supplied: 12 kg/cm2 g (10.5-12.5 kg/cm2 g)
Steam received: 11 kg/cm2 g (9.5 - 12.5 kg/cm2 g)
2. Steam supplied: 3.8 kg/cm2 g (3.5 - 4.3 kg/cm2 g)
Steam received: 3.5 kg/cm2 g (3.2 - 4.3 kg/cm2 g)

Edited by kkala, 10 April 2012 - 10:55 AM.

[Technical Bard]

Hi,
Having no superheat could lead to condensation. Usually, designers choose at least 10 to 20 degC superheat in the network to avoid condensation.

I would argue this is pointless in most places. Superheat should be provided if you are running steam to a turbine, where superheat is necessary to limit condensate in the low pressure stages of the machinery. Low superheats are almost pointless, because the enthalpy in the superheat will almost certainly be consumed by heat losses in the piping. I had a client demand 15°C superheat in LP steam lines, and we did detailed heat transfer calcs showing that even with 2" insulation on steam lines, that superheat disappears in just a couple of hundred meters of piping, leading to condensation.

Steam headers, even when you do have superheat, still need to have traps to collect condensate and remove it from the process - failure to do so will result in slugs of condensate flowing down the line.

[kkala]

Having observed traps even on steam lines of 40 Barg - 400 oC (superheat=148 oC), I thought they were useful during start up, when the line was cold. A past thread concerning transfer line for same steam quality ( http://www.cheresou...86-steam-traps ) seemed to recommend traps every approx 30-60 m of the line. These are not expected to collect condensate in normal operation; or in e.g. tees, elbows, valves, condensate formation is possible?
If that 40 Barg steam were used for heating (intended to be saturated at consumption), what temperature would be recommended at the start of the pipe? Finally what superheat was applied to those LP steam lines, the owner of which demanded 15 oC superheat?
Info like the above would be useful for the subject of superheat, since varying opinions may occur. Topic needs some discussion to be assessed and clarified. See http://www.spiraxsa...ing-theory.asp , text just below diagram at the beginning.

Note: Referring to post No 8, design pressures of mentioned steam lines are 13.0 barg and 5.3 barg (for both suppliers and receivers).

Edited by kkala, 12 April 2012 - 02:25 PM.

[Technical Bard]

kkala,

In the end, our LP steam (4 barg) was not supplied with any superheat. Saturation was acceptable, and traps were provided to collect losses. The steam balance in design needs to account for these steam losses, which we estimated at 2% of total steam flow.
I have recently worked on facilities moving high pressure (100 barg) saturated steam over distances of up to 5 kilometres. For these, traps are provided every 50 metres, as well as at low points in the routing and near elbows. Expectations are losses of up to 5% of the steam to condensation in the line.

[kkala]

Thank you, Technical Bard, for the above data. I had considered some superheat as necessary, even if steam is exclusively used for heating; now there is an option (applied in the field), where superheating can be avoided. 4 barg steam would require an estimated superheat of 18 oC to balance 2% condensation, plus some margin. 100 barg steam would need ~ 11 oC superheat to balance 5% condensation, plus margin.
Can such superheat totally prevent condensation in normal conditions? I do not have experience, advice welcomed, probably limited local condensation occurs.

[Art Montemayor]

My field experience agrees entirely with Technical Bard’s pragmatic approach to an engineering solution. In my opinion, superheated steam should be a mandate when it is a process requirement – such as a steam turbine designed for superheated steam feed. Of course, I have also operated many steam turbines on saturated steam fed by saturated steam headers.

Under some very unique and specific situations superheated steam might be justified for transport and heat loss purposes – but I have never applied it in 52 years. I have, however, had to work with superheated steam headers supplying process heaters and it has been a mess dealing with the pre-requisite de-superheaters and their operation and maintenance. I would always opt for a saturated steam supply header when feeding process steam heaters.

As Technical Bard points out, there is always a design requirement for strategic low point condensate drains and steam traps within a steam distribution system – whether the supply is superheated or saturated. Condensate formation in the distribution system is always expected in any steam header installation. Agreeably, more header condensate means a less efficient operation; however, the installation and maintenance of a superheated system also involves additional capital and operational costs. I always tend to side with simplicity, practicality, and common sense – and a saturated steam supply meets those objectives when applied to process heating.

The OP of this thread does not specify what the proposed steam system is feeding; but at 6 barg, I believe that I can safely guess that it is dedicated to process heating and not efficient steam turbine drives. To reply to the specific query, I would opt for a saturated steam distribution system – adequately and economically insulated, of course – operating at a pressure level that ensures that the target saturated pressure conditions are met by employing the necessary header diameters for the header length required.

[kkala]

Thanks, Art, for the post on steam superheat. So there is the option to specify steam boilers without superheaters, if the steam generated is used exclusively for heating. This brings another issue, also touched in the present thread. Has superheated steam a disadvantage compared to saturated steam, when used for heating?
We would generally assume so, and try to desuperheat steam upstream of heat exchanger. It seems logical, since partial heat transfer coefficient of condensing steam is higher than of a dry gas. But is it clearly written somewhere? Is it supported by experience?
α. Once (1978) we concentrated phosphoric acid in a single stage evaporator, using saturated steam. Actually steam was superheated, we made it saturated by inserting water upstream into its line. Tired of bang-bang in steam line, I entirely closed the water valve. Evaporator capacity was not reduced, probably it went higher. But industrial instruments did not allow precise conclusions.
β. McAdams (Heat transmission, McGraw-Hill - 1954) reports that heat transfer coefficient of superheated steam is lower than that of saturated steam, but the difference is more or less balanced by the higher temperature. I do not have this book at hand now, but I remember a search related to an LPG vaporiser (1984).
γ. Later I searched the issue in the "Efficient use of steam" by Oliver Lyle, yet no clear view was detected on the subject.
So the question is whether desuperheating is worth while or not. It may be "yes" for high superheat and "no" for low one.
Clarifications on this topic (hopefully not advising to calculate heat transfer coefficients!) from members having relevant experience would be appreciated.

Edited by kkala, 14 April 2012 - 11:33 AM.

[Dacs]

This brings another issue, also touched in the present thread. Has superheated steam a disadvantage compared to saturated steam, when used for heating?

Not heating per se, but there are applications that would want to avoid localized hot spots in heat exchangers that a superheated steam may bring, thereby wanting a saturated steam feed.

As a side note, one fellow that I worked with mentioned that in his refinery, they removed a desuperheater upstream of a heat exchanger (due to issues with it) and it did not negatively affect the performance of the heat exchanger.

[fallah]

As a side note, one fellow that I worked with mentioned that in his refinery, they removed a desuperheater upstream of a heat exchanger (due to issues with it) and it did not negatively affect the performance of the heat exchanger.

Dacs,

If it had already been designed to transfer latent heat of saturated steam, it might be said that:

1) Degree of superheating for mentioned steam hasn't been so high.
2) Excess heat transfer area of the HX has been adequate to remove sensible heat of superheated steam till its saturation point and has been acted as desuperheating section of the HX and the rest to remove latent heat of saturated steam.

Fallah

Edited by fallah, 16 April 2012 - 06:50 AM.

[kkala]

Thanks, Dacs, for the valuable information. It complies with cases mentioned in post 14. So apart from higher temperature (yes, it could be a drawback) superheated steam has not a disadvantage over saturated steam? And practically same partial heat transfer coefficient can apply to either saturated or superheated steam? * The opposite concept is widely accepted, at least locally; so additional evidence from other members would be useful.

* Considering saturation temperature as temperature of superheated steam.

[acer_asd]

Dear community members

First of all, i would like to thank you all for having such a great discussion. Whenever I come here, i find a NEW interesting discussion. I always enjoy the discussion related to process engineering.

Just to add to this discussion, I once simulated a heat exchanger on HTRI exchanger suite which was heating a process stream using the steam. This exchanger was 10% over design. So to understand the effect of superheat, i first used saturated steam and noted down the heat duty and outlet condensate temperature. In this case, condensate was sub-cooled. Then i gradually increased the temperature of steam and noted down the heat duty and condensate temperature. I observed slight increase in heat duty first and then started decreasing and return condensate temperature started increasing and after certain point steam was not condensed.

The reason for this was exactly same as mentioned by fallah in reason 2. So you can remove desuperheater if you have some margins in heat exchanger area.