water injection in gas turbines 1

[tbwttihs]

Does anyone know of any good sources of information on this topic? I am pretty much starting from zero here, and want to get a general idea of what it's all about before I start on the bulk of this undergraduate project I'm about to embark on. Any information about texts, online information etc. that anyone knows of would be gratefully recieved.

thanks.

c.

 

[hacksaw]

try any of the g.t manuf.

it is to cool the inlet air

 

[mbeychok]

Craig:

I'm sorry to have to contradict Hacksaw, but the basic reason for water injection into gas turbines is to reduce the formation and emission of nitrogen oxides (NOx). If you do a Google search on "gas turbines", "water injection" and "NOx", you will find a plethora of information on this subject.

Milton Beychok
mbeychok@xxx.net (replace xxx with cox)
(Visit me at www.air-dispersion.com)

 

[timbones]

Actually, depending on the context, both of the above answers are right!

Water injection is sometimes used for interstage power augmentation on gas-turbine compressors. The idea is to add water between stages of the compressor to cool the air and reduce the power consumed in compression. Its usually used only on hot summer days for performance improvement. For information on this check out the "Sprint" version of the GE LM6000 line of gas-turbines.

The other common use for water injection is for NOx control as Milton described. Water is injected directly into the combustor through a modified injector. The principle is that the water reduces the temperature in the combustor enough that the formation of NOx is reduced.

To study this in any great depth you will need to become fairly familiar with combustion kinetics and the different methods for NOx formation. Try a search on prompt NOx and thermal NOx. You will find that to keep the NOx produced off the engine to a minimum takes a great deal of careful control of combustor temperatures. Too hot and the diatomic oxygen in the combustor tends to dissociate and attack nitrogen to form NOx. Too cool and there isn't enough heat to complete the oxidation of CO to CO2. Start with a good combustion text.

I think that water injection is starting to fall out of favour for NOx control these days. The water needs to be extremely pure and fairly large quantites are required. It's expensive and the environmentalists don't like developers using so much water when it just dissappears into the atmosphere. I think the industry standard for NOx control is now Dry Low NOx (DLN) or Dry Low Emission (DLE) combustors. Every manufacturer has a different way of doing it but they all consist of some sort of staged combustor to prevent hot spots.

tim

 

[hacksaw]

team work!

 

[jay165]

Another reason (and the primary one, IMO) to inject water is to clean the turbine blades on-line and thus increase the time between shutdowns and solvent soaks. You need deionized water to do this to avoid SCC of the blades. I've seen a presentation by a company that makes deionization units for this purpose and they claim 3-4 months payout in fuel savings and reduced off line cleaning due to cleaner blades. Website:

http://www.dectron.com

BTW, besides cooling the inlet air, the water increases the turbine power because the mass flow through the turbine is also inceased. More mass flow = more power. The early B-52s used MeOH injection on takeoff for power increase.

 

[davefitz]

Do a search on the GE Lm 6000 "sprint" gas turbine for an update on how spraying water at an interstage of a compressor will boost output.

Likewise, for the impact of water spray at the inlet , do a search on "Mee fogger". For gas turbines installed in dot dry climates, the inlet spray can boost outp0ut by 15%, and up to 20% if overspray to "wet compression" is realized.

For a standard gas turbine without spray or intercooling, the compressor absorbs 66% of the power output of the turbine, and the net output to the generator is only about 34% of the turbine power. If you spray or intercool, the specific volume of teh air being comrpessed decreases, which reduces the paristic power loss of compression plus increases teh flow passing capability of the compressor.

Other issues must be addressed if you spray. The particle size fot eh spray should be less than 200 micron to avoid erosion of the first row of compressor blades, and the water should be deminrealized to avoid accumulation of deposits. Some gas turbine vendors will void your warranty if you overspray to wet compression.

 

[tbwttihs]

thanks for all the info and advice guys. i now have a few more details on this project i'm undertaking. it will be based on this (

http://www.cussons.co.uk/en/products/p9005.html)

dinky little educational GT set up we have in the lab. It consists of an old holset turbocharger unit to act as the gas generator compressor and turbine, a combustion chamber and a free power turbine connected to a generator. It runs on bottled propane gas. From a report of a predecessor, who had the task of trying to get this thing back to an operational state from that of disrepair a few years ago (with mixed success), i have gleamed the a few bits of performance data which may or may not be entirely accurate - namely the operating temperatures and pressures at various stages and the flow rate of air and fuel.

I am currently trying to get this thing up and running again, and get comprehensive performance data. then i want to investigate the effects of water injection into the combustion chamber, so will have to modify the rig accordingly.

what i essentially would like to know from anyone who cares to help out here, is firstly, for a ball park sort of figure as to what sort of water flow rate would be typical in this application, relative to air or fuel flow rates (mainly to give me an idea of how i'm going to practically make the modifications). secondly, and more importantly i suppose, so far i have found little in the way of the actual science behind this practice and would like any help i can get on how i should go about theoretically modelling this whole thing. should i just go about it like i would the standard cycle, and make the relevant modifications to mass flow and specific heat terms?

i guess what i'm asking for is a kick in the ass in the right direction as to what all my equations should look like, or at least where they should be coming from.

cheers.

 

[davefitz]

I think the first step is to know what are your objectives. The choices are:
a) lower NOx emissions
b) power augmentation
c) efficiency improvement
d) none of the above, but get a passing grade in a lab course

 

[tbwttihs]

yes, that would probably be options (b) and (c) then.

 

[davefitz]

Usually the way that water is used to improve output and efficiency are the following:

a) reduce the temperature of the air being compressed. This involves sparying a mist of water at the compressor inlet or at an interstage location, but not in the combustor.

b) steam augmentation. The water is pumpt ot sitable pressures, circulated in a heat recover boiler that is recovering heat from the exhaust gas of the turbine. Inject the steam into the combustor to augment the power output.

c) replacement of tempering air with flash steam , so you can lower excess O2 to low levels ( 2% O2 in exhaust). This involves pressurizing and heating water to about 700F ( at pressures above 3000 psig), injecting/ spraying the saturated liquid into the combustor and using the resulting steam to temper the flame adiabatic temp down to a level suitable for the turibne inlet ( 2400F for modern turbines). I am not sure this has been demonstrated yet.

All of the above require use of demineralized water.

 

[btrueblood]

As far as analytical procedures, get ahold of _Aerothermodynamics of Gas Turbine Engines_, and similar books by Gordon Oates. Took several courses from this gentleman, and his explanations of how to conduct "first principles" as well as more detailed analyses of various g.t. cycles was priceless.

 

[timbones]

If you are injecting water directly into the combustor, the primary objective should be NOx control. It does provide power augmentation as well, but there are much easier, simpler and less mainenance intensive places to spray water if that is what you are after.

The power augmentation has nothing to do with heat of compression and saving power on the compressor if you are injecting directly into the combustor (at this point you are already downstream of the compressor). Power augmentation is caused by the mass flow of the water vapour going through the turbine stages. An efficiency improvement is realized because you don't have to compress the water to the combustor pressure, it is all done with a pump which is much less power intensive.

Definately monitor the NOx on your engine as well as the combustor temperature (usually an infered value from the exit temperature of one of the turbine rows).

It has been a while since I worked out how much water was required to get effective NOx control but I do recall that the quantity was significant (like at least 20% of the fuel mass flow rate). Make sure that it mixes well in the combustor. The last thing you want is high velocity water droplets flying into your combustor liner or first stage turbine.

One very good text on the subject of gas-turbine design is "Gas Turbine Theory" by Herb Saravanamuttoo.

 

[mechanical13]

Hello,

This in reply to davefitz's posting. Inlet "fogging" or Misting has, as a matter of fact been used today. One example is SwirlFlash Technology (Do a google search). They claim to be able to increase power by upto 10% and reduce emissions by upto 40% for non DLN and 25% for DLN. They also have exhaustive theory to support their product.

 

[EdStainless]

Just a couple of practical comments. You will need to use very clean water (resistance breater than 25 megaOhms) and very fine atomizing injectors.

If you can get access to the ASME IGTI you can get tons of info.

http://www.asme.org/igti/

As an aside, one other use for water in GTs is to steam cool the first stage turbine blades.

= = = = = = = = = = = = = = = = = = = =
Corrosion never sleeps, but it can be managed.

http://www.trenttube.com/Trent/tech_form.htm

 

[onebug]

davefitz (and others),

you mentioned that the droplets should be less than 200 microns to avoid erosion of compressor blades. i am interested in looking into this. could you provide some references, papers, etc. on this topic? thanks.

 

[jsummerfield]

From marketing literature, GE offer water injection, steam injection and Dry Low Emission options on several turbines for engine mounted NOx abatement. These include the LM1600, 2500 and 6000 in all of the arrangements.

“The LM6000 Sprint™ System is composed of atomized water injection at both LPC and HPC inlet plenums. This is accomplished by using a high-pressure compressor, eighth-stage bleed air to feed two air manifolds, water-injection manifolds, and sets of spray nozzles, where the water droplets are sufficiently atomized before injection at both LPC and HPC inlet plenums.”

“The introduction of steam or water into the combustion system:
_ Reduces NOx production rate
_ Impacts the gas turbine performance
_ Increases other emissions, such as CO and UHC
_ Increases combustion system dynamic activity which impacts flame stability
_ The last item results in a practical limitation on the amount of steam or water which can be used for NOx suppression.”

“GE’s current, guaranteed minimum NOx emission levels for various control options. With steam or water-injection and single fuel natural gas, the LM2500 can guarantee NOx emissions as low as 15 ppm. For applications requiring even lower NOx levels, other means, such as selective catalytic reduction (SCR), must be used.”

Also search Solar Turbines for such information or details specific to your interest.

John

 

[davefitz]

The 200 micron was a typo- the correct value was that the most of the fogger water must be less than a 40 micron size. This generally requires a nozzle pressure drop of over 1700 psid. There are several papers ( IJPGC and powergen conferences circa 1996-2001) that describe this , by the various fogger manufacturers ( try Mee industries, Monrovia Ca). There are also various ways to descrobe the size distribution, and simply defining the mean or median size is not adequate with foggers.

The purpose of teh small particle size is to avoid erosio of the blade ; small particles , which have a very low saltation velcocity , follow " the law of the wall" and do not impinge the wall surface. However, regardless of how small the particales are at the nozzle exit, there is always agglomerization of the mist by its impaction on the duct stiffeners and bellmouth, so large particles will enter the gas turbine .

 

[abeltio]

wow.
there is a lot of info.
Just for the sake of clarity, i will summarize how, where and why water is "injected" in the GT (Industrial Type and not LM - aeroderivative units, as mentioned in the postings above)
In all cases DEMINERALIZED WATER IS NEEDED.

1. Inlet Foggers or Evaporative coolers:
After the air inlet filters, but upstream of the silencing section.
Used to increase power output in high temperature and DRY environments.
N.B. in high ambient humidity locations: chillers are recommended, the excess humidity condenses (actually A LOT of water).
This solution increases output at the cost of some heat rate increase.

2. On-Line Compressor washing
This water is sprayed into the compressor inlet and should be used for compressor washing ONLY.
Using this water for power augmentation is strongly advised against as there is no control of the flow.
there are limitations on the use of on-line-compressor-washing due to low ambients and during inlet heating operation (compressor extraction) to avoid icing in the compressor.

3. NOx control (injected at the fuel nozzle, i.e. before burning the fuel)
This is a REAL INJECTION.
The previous two are not really considered an "injection", because the water is introduced at atmospheric (at the inlet) or sprayed at very low pressure (~50 psig) - compressor washing.
Water injection for NOx control in modern industrial type GT's is used with liquid fuel. When using GAS only as fuel, most manufacturers developed the DRY NOx control.
WET NOx control (with water or steam injection) has a limitation in the amount of NOx abatement attainable, compatible with flame stability (about 42 ppm).

The amount of water injected is calculated as a function of the FUEL injected and cannot exceed a certain percentage of the mass air flow.
This quantity is biased to be above the compliance level and the humidity in the combustion air is considered as a credit in the calculation (i.e. the amount of water in the air is deducted from the amount of water to inject).

Care should be exercised to during the start and stop of the system to avoid blowing out the flame during initiation and to avoid exhaust temperature overshoot if the system stops suddenly.

4. Water injection for Power augmentation.
Not used very often.
Operators usually prefer steam injection for power augmentation.
The reason is that there is a limitation in the amount of water that can be mixed with the fuel before ignition without endangering flame stability.
The Steam injected for Power Augmentation is injected AFTER the combustion (also called CASING STEAM).
As there is no flame to blowout, the amount of steam that can be injected is far higher than water for power augmentation. So is the increase in power.

Any of the 4 different ways to introduce water in the GT will cause:
output increase.
heat rate increase.

the water that is introduced with the air into the compressor is NOT CONTROLLED (#1 and #2 above).
in case of the water for washing: it only washes the COMPRESSOR SECTION not the TURBINE SECTION.
The highest efficiency and output gains are obtained when operating with a clean compressor. Remember that the compressor takes about 2/3 of the thermal energy released in the unit.

HTH







saludos.
a.

 

[gtsim]

As stated above water injection reduces NOx but increases CO and UHC because of the chilling of flame. Also, water injection increases fuel consumption and thus increases heat rate. Water injection can also be used for power augmentation. Such power augmentation also has an adverse effect on turbine creep life usage.

We have developed a couple of gas turbine simulators, which illustrate the effect of water injection of GT emissions and performance. It also includes a free e-book and discusses all these issues and used with the simulator. If interested you may visit

http://www.gpal.co.uk

and download it for evaluation.