Desuperheater Control 4

[AllisChalmers]

We're using a Schutte and Koerting 6985C as a desuperheater. Also have a temperature transmitter down the line about 30'-0" per their request. Trying to control steam to saturation temperature by adding water. Initial thoughts were led to believe to control by pressure since manual says that pressures should be equal. However, after being out at start-up, we think it should be controlled by temperature. However, it is real hard to get a PID loop to get fine tune control. Any suggestions ?

 

[athomas236]

Its an interesting problem, usually folks avoid trying to control to saturation temperatures because of the obvious problems. A usual approach is to control on temerature to give 25C margin on the saturation temperature.

athomas236

 

[rmw]

See thread 124-84220 for a post dated Jan 17, 2004 for an explanation of how to control at saturation, as well as other posts that relate to your question.

 

[wde3477]

We use a desuperheater to control steam sootblowers on some emission equipment. We ended up controlling the conditioned steam's pressure using the supply steam valve, and controlling the conditioned steam's temperature with the attemporator (water) valve, both with standard PID controls. The trick was with the limited instrumentation available and the interaction between the pressure loop and the temperature loop. The attemporator controls ended up with a derivative gain factor (which became very tricky to tune for the relatively fast time response loop), and an adaptive gain based on the position of the steam supply valve - basically trying to estimate the steam flow rate as a means to set the proportional gain of the attemporator controls. We try to maintain a target of about 300 degrees superheat, and can do so across varying steam flow demands within +/- 10 degrees but this was not without a lot of effort - autotuners just could not develop gains that were robust across all conditions, so it became a manual tuning approach.

 

[abcmex]

from S&K (

http://www.s-k.com/pr_dsh.htm)

"The Figure 6985 is an annular venturi desuperheater. Superheated steam is directed by the cone into the annular area between the cone and pipe wall, increasing both velocity and turbulence.
Cooling water is introduced through a narrow slot (or small jets in the 1" and 1 1/2 " sizes) in the cone at the point of maximum velocity. The combination of velocity and turbulence improves atomization and produces maximum desuperheater effectiveness. The water pressure required should equal the operating steam pressure "

I suppose they want to tell you that the needed water pressure will be around the steam pressure. Other desuperheater configurations require a higher water pressure. It would be advisable to check with the supplier.
Probably the desuperheater should control the water amount using the steam temperature as target. If your steam pressure is very variable you may have problems to tune in.
RGS

 

[rmw]

Sorry to be an old wet blanket, but as a general rule, good desuperheating is dependent on the atomization of the water into fine droplets that can be evaporated, and cool the steam in the process. The better the atomization, the better the desuperheating process.

Good atomization takes energy, and sometimes, lots of it. I am highly suspicious, notwithstanding the designs of particular brands, of desuperheaters from whomever that don't have the water pressure at some value well above line pressure, unless they have some kind of mechanical or physical "wizard" that does the atomization that normally requires the higher pressure. Copes vulcan had or has, I don't know any more, their VO 76 which had a floating plug in an anular orifice, and that thing would get good atomization in the high velocity zone as the steam flowed around the plug, so physical features made up for lack of water pressure, and these things would control as close as 5F from saturation with ease, and with lots and lots of turndown.

Turndown is also a factor in selection of the style of desuperheater you choose. The style mentioned, in a quick glance looks like it would not be good for high turndown. If you need good control, and good turndown, most manufacturers have a steam atomized model, where the lack of high water pressure is made up by the energy in the atomizing steam that "blasts" the water into smytherines.

I was also suspicious of the 30 ft downstream distance to the temperature probe, and more suspicious when I saw the type of desuperheater that it is. The farther you can put the temperature probe, the better the mixing, and temperature equilization.

In this model of desuperheater, it appears to me that high velocity would be essential to good atomization and mixing, but detrimental to a temperature probe only 30 ft from the DSH. Just my opinion.

We weren't told how close A/C wanted to control, or what his "from and to" were, so it is hard to evaluate this application.

 

[davefitz]

general rules of thumb for desuperheaters:

a) there should be a minimum straight downsteram pipelenght to provide a 0.15 sec minimum steam residence time, between spray nozsle and first downstream bend

b) the downstream thermocouple should be at least 0.3 sec downstream to ensure accurate reading of outlet temperature.

c) outlet temperature control logic is normally limited to desuperheat to not less than 20 F superheat, assuming items (a) and (b) are provided.

d) if outlet temperature control logic is used, then there should be an override to prevent water hammer from occurring during low load/ low turndown cases. A good override for some applications is that the ratio of spray water to total outlet steam flow should not be permitted to exceed the ratio of 0.25 :1. This implies that a water flow element be provided and there be some means of estimating the total downstream steam flow.

e) spray water evaporation is improved by fine atomization plyus using the highest temperature spray water available. Best spray atomization turndown is provided with the spring loaded nozzles.

f) an alternate control logic is to use enthalpy control and not temperature control.

 

[GenB]

Good equipment is the answer,
look at KOMAX rancho Dominguez, Ca. they make the best. their info website can help you./
ER

 

[hacksaw]

You can never achieve ideal saturation. Even with the best desuperheating valves about the best that you can achieve is 2-5 Deg C.

Generally if you put the setpoint of you temperature controller at the sat. temperature, it will be unstable.

Comments about location of the temperature sensor are good but optimistic.

Typically you need 50-100 pipe diameters and a couple of elbows...

 

[rmw]

Hacksaw, in thread 124-84220 on Jan 17, 2004, I described a method of controlling at or anywhere around saturation that you want to, assuming, of course you have a place to pressure reduce your desuperheated steam to, in order to accomplish an adiabatic expansion to resuperheat it at a lower pressure.

This system was installed industrially, in the Pulp & Paper industry, at the paper dryer rolls, and so it is proven. It does, however, take a good engineer to understand and operate it, as it is counter-intuitive to non technical people.

 

[hacksaw]

rmw,

no disagreement on the process you are describing; my comments refer to traditional steam temperature control. the latter are not compatible with setpoint control at saturation. the reason being that once saturation is achieved there is no longer a single valued relationship between steam temperature and the water demand for desuperheating.

we've used humidity control in various process applications, though not necessarily with steam.

 

[rmw]

Hacksaw,

We have no disagreement in the comments that you had made. I often characterize the problem in simplistic terms in that as the control variable passes from superheat into saturation, the controller gets lost, because at and below saturation, the temperature for a given pressure is the same, but the quality is indeterminable. Hence the controller is "lost"...it doesn't know where it is, and which way to go to get back to, or above saturation. Kind like being in a fog. Many plant people I have to describe this condition to are not technically savvy enough to use the words you used. Right over the top of their heads.

 

[hacksaw]

the difference is constraint versus setpoint control.

in constraint control, the process gain changes discontinuously at the control boundary, in this case the point of saturation.