Water Quality for IST OTSG

[seo]

We are commisioning a 2x1 combined cycle cogeneration plant. 2 LM6000PD's; 2 IST OTSG's ; 1 siemens Demag Delaval STG.

Sinct the boiler is a once through, the limit on cation conductivity is 0.25 microSiemens. Our demineralizer is producing 0.6 microSimens specific conductivity. We have 304 SS piping from the demin to a 304ss Demin Storage Tank. Makeup is to a 304SS FWHT via 304ss piping from the Demin Storage tank to the FWHT. The feedwater system to the OTSG is 304SS as well. The piping form the condenser to the FWHT is CS.

We are having problems with conduxtivity in the demin tank. In a few days the conductivity in the tank will rise to 0.6 microSiemens.

We have inspected the tank and found ehat apperas to be rust on the bottom brought in from the inlet piping. Tank has been cleaned and repassivated. Conductuvuty cintinues to climb. Inspection of the pipe reveals some rust on the welds. Not sure if this is from incorrect welding rod or just collection of rust where it hits the weld bead. Pipe is ATM A312, but has not been passivated.

We have been talking to other plants with similar configurations. Some have put final bead filter/ion exchange right at the connection point for the feedwater to the OTSG. This is an expensive option. I am not sure we will ever get the conductivity down to 0.25 w/o it.

Any suggestions or insight are appreciated.

 

[MJCronin]

seo..

I think that you should consider reposting your question in a forum more devoted to materials selection and/or steam power plant operations and design:

Consider:

http://www.steamforum.com

http://www.steamplantengineering.com/

and

http://www.hghouston.com

Please come back and tell us about your preferred fix and the reasons for it

regards

MJC

 

[cub3bead]

The conductivity increase is due to absorption of atmospheric CO2 by the demin water while in storage. The only way to stop it is to put a N2 blanket on the storage tank.

My understanding of IST's feedwater requirements has lead me to conclude that the only way to comply is to have a feedwater polishing system, either powered resin or mixed bed. operating without one will surely result in voiding of the warranty.

You will need the N2 blanket even with polishers to extend times between regenerations or pre-coating because the CO2 absorbed in unblanketed tanks will exhaust the anion resin prematurely.

The rusting you see in the tank and piping maybe due to workers using tools (grinders, wire brushes, etc.) that have been contaminated with ferrous iron or made from ferrous iron. Tools for stainless stell fabrication must be segregated from those on carbon steel.

 

[davefitz]

cub3bead is right , in that the increased conductivy is probably CO2 , but one also needs to consider the possibility that there is a condenser leak. In your statement that the tank conductivity increases over time, is that under the condition of zero influx from the condensate pump?

Although the CO2 will increase the conductivity, it might not be deleterious to the cycle , and might be permitted by IST. There is another means of measuring conductivity ( de-gassed?)which eliminates the CO2 effect and focuses only on the ions that will damage downstream equipment , and perhaps that conductivity is the governing effect. We have the same issues on drum type boilers, wherein the breakdown of organic amines will result in a high conductivity of the steam due to CO2 formation; to meet steam turbine specs, we would use a degassed (?) conductivity reading to judge acceptable steam conditions.

If the IST unit is their typical design, they have tiny orifices at the inlet of the economizer ( for flow stability) , and they cannot have any traces of iron oxide in the feedwater to avoid pluggage of the orifices. Your comments of using a CS line from condenser to FWHT may be a source of the oxides, espescially if you are using traditional AVT alkalien feeedwater treatment methods ( ie, it forms magnetite , which easily spalls off). I would either replace the CS line, or consider using the combined oxygenated feedwater treatment method.

As far as needing a polsihing system, Alstom had proved with their once-thru HRSG's that a polisher is only needed for the first month of operation, assuming the condenser is tubed with stainless steel tubes. After the first month , the cycle is cleaned up and the rented polsiher is returned to the rental agency. Historic expereince with conventional once thru units , on Rankine cycles with regenerative feedwater heaters and copper tubed condensers , does not strictly apply to the case of a once thru HRSG with a stainless tubes condenser.

 

[cub3bead]

Did you look at the fine print in the IST warranty? You will likely void the warranty if do not polish the feedwater.

With 0.6 uS/cm coming out of your DI system you need either a polishing strong acid cation or a mixed bed exchanger to make specified water.

 

[RTH2Q]

No offence but I would disregard the last statments made on the post by davefitz. Your MBs should be producing water quality of around 0.15 uS/cm of specific conductance at a pH of around 6.8-7.0. True you will see an increase in spec.cond. due to dissolved gases ie: CO2 and O2. This effect is minimized with a N2 blanket over your demin tank. Your contamination is likely due to ferrous by-products forming in your condensate return due to the CS material and the fact that even after this tank is drained flash corrosion occurs on wetted parts the only solution is to clean and passivate the tank. A cation polisher will clean up this condensate while you are commissioning. Once your plant operation is stabilized your condensate return will become 90 % of your make up to the OTSG and the effect of the Demin tank high TDS will be minimized. PS. Amines break down to form Ammonia compounds not C02. CO2 is formed when CarboHydrazine breaks down at high temps. Hydrazine does not break down as readily but most facilities avoid this product due the EH&S concerns. In addition, you probably wont need O2 scav. for the OTSG when commissioning. CO2 is usaully formed in cycling boiler water systems due to the break down of Alalinity at high temps. ie: in HRSGs. Hope this information helps.

 

[seo]

An update.

I had some errors in the post. The mixed beds produce 0.06 to 0.1 microsiemens conductivity.

We did clean the demin tank and re passivate it.

We have come to the same conclusion about the CO2. We are looking at polystyrene balls to reduce the surface area of the tank and limit CO2 absoprtion.

We have installed a degassed cation conductivity analyzer on the feedwater tank.

Our mixed bed demineralizer is sized for 750 gpm. 17 GPM is for the makeup to the cycle, and the rest of the capacity is for polishing condensate return from the steam host. We have also installed a bypass fron the discharge of the condensate pumps to the demin. This will polish the condensate during times of low process steam load.

I think these changes will get us where we need to be.

Since some of you have experience with IST units, we are experiencing strange pressure oscillations at part loads on the units. This occurs even if we drop the feedwater valve into manual. IST is looking into it, but have not come back with anythiong yet.

Anyone else have this issue? I suspect that the the flows to the tubes is not uniform.

 

[davefitz]

You are right about the flow to the tubes not being uniform. That is why they add small orifices to the inlet of the boiler- the orifices are sized to cause the pressure drop thru the orifice to be about 400 psid ( 27 bar) at 100% load, which means the pressure drop thru the tiny orifices is only 36 psid ( 2.5 bar) at 30% MCR load. That is not enough pressure drop to prevent the Ledineg static flow instability from occurring when the boiler is configured as a single Mother-of -all passes without intervening mix headers. The problem is related to the drastic increase in fluid specific volume when the liquid is flashed to steam in the evaporator section.

IST has also lately licensed a different boiler technology ( Benson from Siemens KWU), which uses an alternative approach- instead of adding 400psid pressure loss in inlet orifices, they use intermediate mix heders at strategic locations to eliminate the Ledineg static flow instability issue. But it lookslike you are using the IST classic technology, so you may need to find a way to live with it.

 

[davefitz]

To be a little more helpful, the startup flow instability can be minimized by starting up with a maximum pressre in the boiler- this reduces the change in specific volume in the evaporator tubes as the liquid flashes into steam.

Usually you cannot artificialy increase the boiler pressure simply by throttling at the steam turbine control valve- that valve is not suitable for such drastic pressure drops and it is instead used for speed control prior to synchronization. Instead, you can provide a startup bypass control valve in parrallel with the boiler outlet stop valve, and size that smaller valve for holding a constant full upstream boiler pressure during startup.

This fix will not completely prevent the flow instability , but it will help .