Corrosion inhibition and sample analysis interpretation of industrial closed loop cooling system

[romke]

I have come across a question about a cooling system where I actually have no clue what the best answer might be.

It concerns a very large (>500 m3) industrial closed loop cooling system that originally was filled with water from a local source 1:1 mixed with a "classic" type automotive antifreeze (NAP-free). After some time in use (meaning about 5 years) there was a major incident necessitating the need for a "top up" of about 50 m3 - which was subsequently carried out by using a modern automotive ready made coolant (1:1 water/glycol with a organic based inhibitor package ("OAT"). A few years later again the company felt the need to have the fluid analyzed, although no specific problems were encountered. The firm that carried out the analysis and the sampling reported that the coolant mixture contained virtually no nitrites and that thus the system was in acute danger of corrosive attack. That firm promptly delivered a nitrite based inhibitor that was consequently mixed with the coolant already in use for years, bringing the nitrite content up to around 1500 ppm - ignoring the fact that both the original antifreeze and the later added ready made automotive coolant did not contain any nitrite on purpose....In the meantime the system performance has gone down a bit to the extend that on hot summer days the cooling capacity can no longer transfer all the heat to the surrounding air, so the underlying amount of processing has to be temporarily reduced.

Since I am fairly familiar with automotive cooling systems and the type of fluids used therein, I was asked what type of testing would be suitable to verify the state of the coolant fluid and if done so, what could be done to correct eventual deficiencies in terms of corrosion protection, scale prevention or freezing protection.

Frankly I have no idea what kind of testing would be appropriate here. Of course the usual physical and chemical tests can be run, but then? You will end up with a bunch of data, but since there is no clue as to what the fresh data were and how these have been influenced by the addition of a totally different coolant and a supplementary inhibitor, there is no way to tell whether the present composition needs any correction, although it most likely does after a few decades of use.

My initial suggestion to drain and inspect the system, clean when needed and fill with a mix of suitable water, glycol and a inhibitor package from a supplier not only knowledgeable on water treatment in this type of cooling systems but also able to analyze samples, understand the results and make sensible suggestions as to what to do when corrections are needed, was not exactly met with much enthusiasm. I can understand that: it would take a lot of time and be very costly, not only because of the cost of water, glycol and inhibitors, but also a very large amount of waste water needs to be discarded.

Question 1: Do you have any suggestions on how to analyze this mix of different fluids and components and how to arrive at conclusions and maybe measures for correction?

They have plans for building two similar plants that will also contain two very large closed loop cooling systems.

Question 2: Has anyone suggestions what kind of inhibitors would then be used best in the water/glycol mix, given the fact that steel and Al protection is needed, that the maximum coolant temperature will be around 70 degrees C and on which regular meaningful checks on coolant composition can be carried out?

Although these systems have a lot of similarity with automotive cooling systems (closed loop), the operating requirements and the sheer size call for a different approach. The idea is that the coolant need not be changed (automotive systems on modern fluids every 5 years), must give very long term protection (decades!) whereas a car or truck usually changes ownership after 5 years or so thus neither the first owner nor the manufacturer worries when corrosion results in the need for repairs, the inhibitor should be substantially cheaper then the pricey "OAT" complexes etc.

Any suggestions you can come up would be highly appreciated.

 

[curtis2004]

You need to talk to glycol suppliers. They usually have tons of information and will guide you through selection process.
As an example Dow Chemicals make Dowtherm (trade name of glycol).
They even sell premixed glycol with corrosion inhibitor added.

Curtis

 

[romke]

Curtis,

I had considered it. And even did talk to one of them. Unfortunately they could not help me with the first question I had: "how to analyze a used sample that consists of a possible mix of individually suitable fluids but not necessarily so when mixed?" I guess most of them will be rather reluctant to comment on these mixtures because the interaction of various inhibitors over time is difficult to predict.

I found a number of answers to the second question, but opinions differ. Most of them could be categorized as "buy my product and you'll be fine", some honestly say "we don't know" and only a few are more clear in the sense that they recommend particular inhibitor types with the restriction that success may well depend on actual operating conditions like temperature, oxygen intake, alkalinity etc. And as said: most of them seem to have limited experience with closed loop systems anyway and if they do usually in a different temperature range.

 

[georgeverghese]

Abstracting from Perry Chem Engg Handbook, 6th edn, page 12-43/44 on MEG - water closed loop systems:

a) Makeup water should be soft, low in chlorides and sulfates - use de ionised water if you can.

b) pH of the recirc coolant should be monitored and maintained at between 8.8 and 9.2

And would also add that the expansion drum be N2 blanketed, and makeup concentrated MEG containers be also N2 blanketed.

These may be found on page 11-95 of the 7th edition also

 

[georgeverghese]

Where do you have aluminum components in this closed loop?

 

[romke]

@georgeverghese:

thanks very much for your reaction

the N2 blanketing might well be a good idea - since it will help against oxygen intake.

as far as water quality is concerned I have some questions. it is stated to be "soft" but the circulating fluid should have a pH between 8.8 and 9.2. to me that seems somewhat contradictory, because I assume "soft water" to have a pH in the range of 6.5 to 8.0. so, where does the additional pH come from, unless you use alkali inhibitors?

I have also seen suggestions that warn against water that is too soft: carbonates present can more or less act as a "natural" inhibitor that covers all internal surfaces when the temperature is high enough for the carbonates to precipitate.

as far as the MEG is concerned: I assume you mean inhibited MEG, since a pure MEG/water mix can be quite corrosive.

we then do have a kind of locally made type of automotive coolant with a more or less complicated mix of unknown inhibitors that will last for a limited period. in vehicles that fluid is simply discarded after a certain time, but actually checking whether it still offers sufficient protection against corrosion is quite another thing, unless the supplier has informed you about what exactly to look for. of course a very low pH value would be cause for concern, but when it stays within the foreseen range it not necessarily is noncorrosive and chemical changes in the inhibitors may well cause additional problems. it is for that reason that I am more interested in a rather simple inhibitor package based on nitrites and maybe a limited amount of azoles.

the aluminum is used in the heat exchangers, a horizontally arranged array of fans and radiators (somewhat like in your car, but then horizontally mounted and much much larger).

 

[georgeverghese]

Think the slightly alkaline pH is to encourage the formation of a passivating oxide layer on pipes and exchanger tubes.

Yes, agreed, we mean the inhibited MEG.

DI soft water can be corrosive if it is oxygen saturated, so if the makeup water inhibitor mix includes an oxygen scavenger, and with the expansion drum is N2 blanketed, that should do.

When you read Perry on this topic, it says that once the inhibited coolant has gone beyond the recommended pH range, there is no point in adding more inhibitor to readjust pH - it recommends a complete changeout.

Presume the aluminum you've mentioned are the external fins on these radiators ? So these arent affected by coolant corrosion, but by corrosive elements in the surrounding ambient air. Standard practice is to use marine grade aluminum for these fins ( series 6000 ?).