Flow divider for Water Injection into Frame 6FA (NOx abatement)

Hi, I would like to understand who has experience with the operation of a mechanical driven flow divider (make Scherzinger) for Water Injection in a Frame 6FA combustion section for NOx abatement. Has anyone experienced the failure of such flow dividers and what was the cause.

I have a good deal of experience with GE-design heavy duty gas turbine water injection systems and components--but I have never seen a driven flow divider used for water injection--and by driven I mean a flow divider that has a motor (probably AC) that is used to provide the pressure/flow required for the application. In fact, I have only seen one or two water injection systems that use flow dividers at all. And, if memory serves me correctly, they were both on turbine packages provided by GE Belfort (France), who is extremely fond of thinking outside of the box (the long-established and proven box, that is). Now, I haven't see EVERY water injection system, but I've seen a fair share of them.

Have you had a read of the Water Injection System Description in the Operations & Maintenance Manuals provided with the turbine and auxiliaries?

It's possible to attach electronic files to threads/responses here at the new Control.com. Can you attach the .pdf of the Water Injection P&ID, please? It would be most helpful; thank you.

If the equipment is having or exhibiting problems, could you describe them?

I did a quick World Wide Web search for Scherzinger flow dividers and couldn't find much information, except for sales blurbs from companies that sell the equipment. (Again--it was a QUICK search.) I would be interested to know if the flow divider is being used for raising the pressure/flow of the water. And, if so, is the "pump" a positive displacement pump or not?

Anyway, if you would provide more information, even just a description of what it is you are wanting help understanding, we can try to provide some assistance.
Hi all,

I have never seen also such configuration ( flow divider) for water steam injection on frame6FA....

Would be great to get more details on type of pump or valves actually used on your unit...

As CSA said ...GEEPE (Ge Belfort) Can sometimes make "original&customized" configuration/arrangement/design for their 6FA units packaging...

Looking to hear more from Pauldh about his plan/unit configuration particularly on that Water/steam injection skid ...

Also did you get a read on that GER 4211 "GE HDGT emissions & control"

If so you can see some notes in this OEM document... "NOx Abatements values" on Water/steam injection... but it does not specify any values for F class/fleet...

Maybe you can highlight us on that point...

Thanks & regards,
Also according to OEM Nox Abatement varying from 6FA.01 TO 6FA.03

Nox (ppmvd ) at base load (@15%O2): 25 for 6FA.01 & 15 for 6FA.03
CO (ppm) at min turndown w/o abatement: 9 for boths 6FA.01& 6FA.03

Thats values are for DLN ... do you know the values on Water/steam injection ...

Also would you tell us which kid of fuel is used ...

Please see the simple design page from MkVIe on the Water Injection system.
Fuel as designed is Dual fuel system with MNQC combustor. Fuels are LPG (1. Propane as one mode fuel and 2. Butane in another mode) and Diesel Oil as liquid fuel to start-up. Unit is indeed designed by Belfort. As mentioned the Flow divider is made Scherzinger Type: 21911MP
GE Part no. 113T2848P001. In the document they mention this device can be used for Fuel and Water injection.
If the flow divider is not driven by a motor or otherwise, then I ask myself how is the speed kept constant at the design speed of this device. CSA mentioned that he had seen flow dividers without a drive, is that correct??
Thanks all for your reply and thinking.



Yes; flow dividers are used almost universally for liquid fuel systems on GE-design heavy duty gas turbines. They do not have to spin at a constant speed--in fact, they spin at whatever speed is the result of the flow through them.

The purpose is to divide the flow EVENLY ()equally) into all of the combustors. I wouldn't be surprised to learn that GE Belfort used a flow divider to also evenly divide water injection flow EVENLY (equally). They have a reputation for doing unusual and unexpected things.

The photo/screenshot you sent was NOT from the Mark VIe--it was from the HMI used to communicate with the Mark VIe. They are NOT the same thing. The HMI can be used with any digital Mark* turbine control system, as well as with GE DCS (Distributed Control System) packages. The HMI is a way to monitor operation, send commands to the unit, and program and configure the Mark V or Mark VI or Mark VIe or the DCS it is being used to control the turbine or a plant. It is also a big alarm annunciator....

When a liquid flows into the flow divider it starts a gear mechanism turning. The gear mechanism is what divides the flow equally into the output lines. This is very important for liquid fuel flow into a GE-design heavy duty gas turbine because equal fuel flow-rates into each combustor mean the temperature of the hot combustion gases produced in each combustor will be nearly equal, and that's good for hot gas path parts life. Very good.

And, it's also important for water injection flow-rates to be as equal as possible also, because water quenches the flame temperature (which reduces NOx formation) and it also affects hot gas path temperatures.

The 6FA GT only has six (6) combustors; GE Belfort might have thought it was "inexpensive" and a way of ensuring the flow-rates of water being introducted ("injected") into each combustor was as uniform and equal as it could be.

SOME GE-design heavy duty gas turbines have a small flow divider mounted on each combustion can cover--because each combustor has multiple nozzles. This is usually for liquid fuel flow.

USUALLY water injection is done through a set of nozzles in the combustion can cover, and the water is often injected not directly into the combustor where the fuel is, but, rather, into the combustor with the air that enters the combustor for mixing with air. But, I haven't seen all combustor designs, and I certainly haven't seen all the designs GE Belfort has come up with.

That's about all I can add. If you are working on a site with this unit, you have ready access to the water injection compartment, and you can see if the flow divider is motor-driven. From the screenshot you provided, it is not, or at least it doesn't appear to be. One of the worst aspects of the GE turbine control system are the HMI screens. They are NOT always good depictions of the P&IDs; sometimes (unfortunately, often) they are often downright wrong.
CSA, thanks very much. So speed of flow divider is related to the pressure in and out. Flow divider sectional drawing is shown below, and the bearings inside are all needle type bearings. These are thus lubricated by the BFW quality type of water that is injected. That does not seem to be a very clever idea at all. In all I count about 19 needle bearings and 2 ball bearings. I have seen a picture of the damaged bearings following the failure of the flow divider.
What would be a good alternative to get away from using the flow divider in the water injection system to the F6FA. Are there any better solutions that indeed work??
flow divider.png

Looks you are no more worry about NOx abatement... values on your unit

But more worry about this Flow divider in the water steam injection...

By the way is that "SYSTEM " also used for both steam/water injection...( just for information)

For sure there is other way for designing such system ( Manifold for example with valves arrangement are used )....


Looks you are no more worry about NOx abatement... values on your unit

But more worry about this Flow divider in the water steam injection...

By the way is that "SYSTEM " also used for both steam/water injection...( just for information)

For sure there is other way for designing such system ( Manifold for example with valves arrangement are used )....

Thanks James. We are worried about the NOx emissions, however the Water Injection system is not reliable.
Thanks James. We are worried about the NOx emissions, however the Water Injection system is not reliable.
Thanks Pauldh .

I will have a better view on this thread ..then come back after speaking around my colleagues about that design from GEEPE Belfort ...

I know some people from GEEPE ( Iam in Belfort right now..) who can advise on alternative solution if youu are not happ/satisfied with that configuration....

Thanks Pauldh .

I will have a better view on this thread ..then come back after speaking around my colleagues about that design from GEEPE Belfort ...

I know some people from GEEPE ( Iam in Belfort right now..) who can advise on alternative solution if youu are not happ/satisfied with that configuration....

Hi James, that would be very much appreciated. What I mainly do not get is that GE has designed a flow divider with about 20 bearings (ball and needle type) in the water without lubrication. That is not going to work, right?
Hi Pauldh,

Thanks for these clarifications....indeed that will not going to work properly in medium/long time operations....

There is according to OEM documents solution on that skid design with drain & control/stop /regulating valve ....installed with manifold only ...

You can try to ask you CPM or GE representative to ask them if they can give you a quote for modifying some elements ...
to get an alternative solution ...

Also have a read on the document GE 4211 ...then you will got a better picture on what iam sayin here...

GE Belfort strikes again!

All the GE water injection systems I have worked on used a single output line from the skid (compartment) to a manifold around the axial compressor casing. High pressure flexible hoses were then used to supply the water to a check valve at each combustion can cover. The water then entered an internal passage in the combustion cover that acted like another manifold and served several spray nozzles (which were really just fuel oil burner nozzles also used in oil-fired furnaces) to provide the atomization of the water spray to achieve the maximum quenching action of the flame temperature to reduce NOx formation.

However—I suspect there’s more to this story and that GE Belfort is using a different method and/or area in the combustor to inject the water and so felt the need to use a flow divider.

I strongly suggest contacting Scherzinger to ask for assistance with this problem. They probably have much more experience with the equipment and the requirements for demineralization water than GE Belfort. You can also ask GE Belfort to be involved; they are usually more than happy to have someone else solve a problem they created—but at least in the past they have shown they learn from their mistakes.

This is a very good example of not working with suppliers and manufacturers to order the proper equipment for the application. Manufacturers and suppliers are very willing and happy to recommend equipment to meet the needs/requirements for a particular application; in this case it might just be a different set of bearings. But all too often OEM engineers just look at a catalog and make a decision without asking a supplier or manufacturer for a recommendation AFTER supplying the requirements/specifications of the application. I’m not saying this is what happened with this issue, and it’s entirely possible the supplier or manufacturer either recommended the wrong equipment or supplied the wrong equipment (which can and has also happened). But, usually, it’s not asking for advice or not fully explaining the requirements.

Please write back to let us know how this is resolved!!! And thanks very much for the information and feedback you have already provided.
Hi every one,
I am intersting in WI divider problem also. My plant also using GE mark 6 FA like Mr Pauldh and have issue with WI. For more information: we using demine water for WI, flow divider for 6 combusion chamber. Some issue:
1. When WI start, exhaulted spread temperature increase more than 20 deg C. (New WI divider do not get that one).
2. After sometimes running gtg with WI then spread temperature increase then gtg need to shutdown. Time to gtg running normal lower than Designed.
3. When we check water pressure at 6 line then it not equal that reason. we do maintenance we found have moods in flow divider gear (flow divider made by carbon steel).
I have 2 possible think:
1. Quality of demine water: control range PH is 6-8, then acidity will made corrosion and dust forming. Suggest chemical dosing to maintain pH higher than 7.
2. Check filter designed is correct to prevent dust attacks to flow dividers.

Anny way, flow divider is excellent ideal to devide equally flow than manifold and it practically give good result in liquid fuel line. So we need consider in different of WI and liquid fuel to achive problem.

Hope to get your expertise support to solve it issue.
Not quite sure I completely understand all of the post, but here goes anyway.

1. This is very indicative of problems with Water Injection flow-rates into individual combustion cans, OR with plugged (choked) injection spray nozzles which can also cause similar problems (and should not be overlooked when troubleshooting spread problems while operating Water Injection). I don't fully understand about the last sentence; I presume it means if you replace the flow divider with a new one the spread problem at Water Injection initialization goes away for a while.

2. Yes; if the flow problem gets worse then the exhaust temperature spreads are going to increase, and reducing load until Water Injection reduces or stops is another method of preventing shutdown--but it doesn't usually resolve the root cause of the spread problem while operating Water Injection.

3. The purpose of the flow divider is (presuming all the Water Injection spray nozzles are clean and flowing properly) is to equalize the flow-rates to the individual combustors which should result in equal pressures in the lines to each combustor which should mean the same amount of water is flowing to each combustor--which means the hot combustion gas temperatures in each combustor should be equally affected by the amount of water flowing into each combustor. I'm not quite sure what "moods in flow divider gear" means, but if the flow divider is made of carbon steel then it's definitely going to be attacked by demineralized water over time.

If you're not using "boiler quality" water for water injection you're not only going to damage and carbon steel components but eventually you're going to cause problems with the water injection spray nozzles and internal hot gas path components. Minerals not removed from the water being injected in the turbine will eventually "plate out" (adhere/stick to) combustion liners, flow sleeves, turbine nozzles and turbine buckets--which IS NOT GOOD. GE provides a fully detailed specification for the demineralized water used for NOx reduction when operating on liquid fuel. This spec should be treated as gospel (meaning it is EXTREMELY IMPORTANT). If you can't obtain or produce demineralized water meeting this specification then you should expect to have problems--some can be VERY catastrophic and result in a long forced outage (down time) and high repair costs--much higher even than simply replacing or sourcing a flow divider made of proper material(s).

"Dust attacks" I'm interpreting as "rust attacks"--iron oxidation. If that's not correct, please clarify. The purpose of filters in the input line to the flow divider is remove dirt, rocks, weld slag, swarf (metal filings from machining operations), cigarette butts and most (but not all!) such solids from reaching the flow divider and probably the flow meter(s) to ensure proper operation of the equipment (flow meters and flow divider with the incoming water flow). (There is usually a strainer upstream of the high-pressure water injection pump to protect the pump from such materials entering the pump suction.) If by "dust" you mean something other than "rust" then the filters should be removing as much "dust" as possible (see the preceding list which is not complete but representative of the types of contaminants which can be entrained in demineralized water.

But I can't stress enough how important the use of quality demineralized water is to reliable long-term operation and availability of a multi-million US dollar asset like a GE-design Frame 6FA heavy duty gas turbine and its auxiliaries. While I'm not surprised by the use of a carbon steel flow divider instead of one designed for use with demineralized water (and this one may be--I'm not familiar with the Scherzinger units being discussed in this thread), it sure doesn't seem like a good idea.

As the Owner/Operator of one of these machines you should be able to contact the supplier/packager (GE Belfort France???) for assistance with this problem. And, there's also NOTHING preventing you from contacting the flow divider manufacturer for information about the suitability of this pump for use with demineralized water in this application (CLEAN, demineralized water meeting the OEM (Original Equipment Manufacturer) specifications!). If they indicate another flow divider would be a better choice, you are free to purchase that device and install it. OR, if the unit is still under warranty from the packager/supplier then you should contact them to work out a suitable replacement for a device which is deemed unsuitable for the application.

As you found this thread with your search engine and see that the prior questioners didn't really respond with any results or information we will ask you to provide feedback as you resolve this issue and let us know what you find and how you resolved the issue. A LOT of people read these threads, not just now, but in the future (this thread is more than one year old) and feedback is most helpful when it's provided to those who read the thread (now and in the future).
Thanks Mr W have brilliant suggest for our flow divider issue. It take time to me to do all of your suggest to find the way. Hope to have positive feedback in futures.
For demine water that used for BFW after de-airrator and injection amine to control pH.
The filter at discharge pump and before flow divider as Mr Pauldh attack.
For dust material is key words need to analysis before we do anything.
If you have anny suggest, please post here or to email [email protected]