any risk associated with a properly executed TSP cleaning?

Binnacle

I've been gradually improving the steam heat (repitching radiators, main and rad vents, low-pressure gauge, Vaporstat) in my recently acquired 1924 row-house.  Thanks and gratitude to HH.

After replacing the nearly opaque level gauge glass I was a bit shocked at the condition of the water: enough rust in suspension that the water was nearly as dark as the old glass.  I drained and refilled the boiler for the second time this season and added two Rectorseal Steamaster tablets and it looks dramatically better, but still plenty-o-rust.  I'm also wondering if there might be traces of oil lingering in the water since the skim-port plug appears never to have been touched.

The next time we have a warm spell I'm contemplating giving the boiler a TSP (trisodium phosphate) cleaning per the Weil McLain instructions:  drain and refill, two cups for the 200k BTU (MBH) and two hours at 180 degrees, then drain and refill again.  I'm afraid to touch the skim port plug mid-season so skimming will have to wait till spring.  Just put in a union below the pressure relief valve for adding the TSP and lowering in a lab thermometer.  I might use the relief valve port to flush/fill the boiler with hot water to minimize thermal stress as water from the make-up valve is winter-cold.

THE QUESTION:  Is there any risk that the TSP cleaning will inspire a leak or any other type of failure?  The reason I ask is that the unit is a 53 year-old gas-fired Weil McLain E-9 manufactured in 1961.  Not exactly ancient, but certainly old enough that some might think it should be replaced.

If a risk of a problem exists I'll put it off till spring to be safe.

Though the Jersey City water company report says the PH is 6.5, the strips say it's somewhere between 7 and  8 and this I suspect is why the unit has endured.  Another factor in favor is that this sturdy boiler has 800 lbs of cast iron compared with the 650 lbs found in the similar power modern EG-65 that would replace it.

************

NOTE: Please see "CRITICAL WARNING" at the end of this thread.

Binnacle

-

I should add that the boiler has a continuous hot-water heater that was clearly active at one time but has been taken out of service in favor of a tank heater.  I'm not sure if this was done to save money in the summer (the boiler is ridiculous huge to run for hot water alone) or if a pin-hole developed in the heater.

ALIGA

my opionion only

i wouldnt touch it with tsp. As long as you dont have foaming, banging, etc you should be fine. Maybe drain all the water and refill a few times.



just let it live.

Binnacle

-

Perhaps you are correct.  I've been reading threads where newer boilers were brought into a state of clear water after cleaning with Steammaster tablets, but I don't think anyone with an older unit was in those conversations.  When I put in the union on the relief valve I noticed the rust on the sides is about 1/4 inch thick.  Could be hazardous to disturb it.

Binnacle

but. . .

On the other hand I see rapidly condensing water in the level gauge glass above the water line--It swirls down the side.  And the level rises and falls in a gentle not-quite-surging fashion unless it is surging and I just don't know what that looks like exactly.  Not "dancing" however.  Subtle sound of marbles jostling when it starts to steam.



Above 8oz there's some water hammer in a far radiator and in general a lot of "swooshing" in the pipes.  But the risers are not yet insulated.



So on the balance I belive there is some oil dissolved in the system.  I do want to skim, but want to leave the skim-port plug for the end of the season.



Haven't seen any stories of disaster from TSP cleaning, so it seems a reasonable idea.  I'm *not* planning on adding caustic acid or lye.

EBEBRATT-Ed

TSP

I don't think the TSP would hurt it but I wouldn't use it unnecessarily. A slight gentle bouncing of the water level is probably not a problem.  if the top of the gage glass is dry and the water level isn't driving the low water cutoff and or water feeder crazy than I would leave well enough alone.

Hap_Hazzard

Gentle Bouncing

Indicates either that there's no serious contamination problem or the gaugecocks may be restricted.



Trisodium phosphate is a stronger base than sodium carbonate or washing soda and therefore intrinsically more dangerous to use. Either can be used to remove hydrocarbon sludge and should be purged from the boiler afterwards, but the risk of damage due to caustic corrosion from trisodium phosphate would be greater, and its use is regulated due to its environmental impact.

Binnacle

-

When I've drained the boiler there was tons of rust but not really any of the black coloration associated with oil.



So perhaps then TSP is overkill, especially as it is unlikely that it will remove the 1/8 to 1/4' inch think coating of rust inside the boiler.  And I suppose that I really do not want to do that!



Unless some folks arrive with unequivocally positive experience and recommendations, I'll content myself with keeping the PH up around 9-10 with the Steammaster tables to minimize future corrosion and draining / flushing the boiler a few more times to reduce the amount rust sediment sitting down on the bottom.



I thought about the water flowing in the glass and it seems it might just be from steam condensing in the copper and bronze fittings at the top and trickling down as water.

Handyman 242

leave it alone

With what you describe happening with the gauge glass it does sound like you have a problem, but TSP can be very strong.  I would keep draining and flushing boiler until the water is clear.   I would then remove the pop safety valve (hold back with another wrench), pipe a couple of nipples and 90s to valve opening,  flood boiler and then skim through pop safety opening.  It may take a couple of hours but with that age on the boiler I would not take a chance with TSP.   I have done this with older boiler where I did not want to disturb the skim tapping and it works 

Binnacle

ok

Thank you for the advice about skimming through the relief valve port.  The idea had crossed my mind but your experience confirms the validity.  With the union in place now on the relief valve I can buy another union and some elbows and nipples and rig-up a skim "arm".  I was planning on opening up the skim port at the end of the season, but perhaps it's best to leave it be.



I like this boiler with its simple Thermopile control system and massive cast-iron construction and want to keep it going as long as possible.  Have already decided to avoid the aluminium block high-efficiency units that are now popular and stay with cast-iron when / if the day arrives that it needs to be replaced.  Even with a subsidy, who needs a unit that will probably fail in less than ten years?



The Rectorseal Steamaster tablets appear to have a mild cleaning effect so a few more flush / fill cycles and a skim or two should make everything nice.

ChrisJ

Aluminum

Who makes an aluminum block steam boiler?

Binnacle

HE cast aluminum

Weil McLain, and all the rest I believe.



http://www.pexsupply.com/Weil-Mclain-383-500-724-Ultra-230-183000-BTU-Output-High-Efficiency-Boiler-Nat-Gas-or-LP



A quick Google will bring up the horror stories.

ChrisJ

Water

Those are for hot water only.

Binnacle

oh

I wasn't paying close attention as I'm not buying.  Just hit Pex to see what the approximate cost would be, though they don't have what I was looking for.  Eventually found a price for a WM EG-65 which is the current equivalent for my old E-9.



If my system was hot-water instead of steam I'd avoid aluminum though.

ALIGA

Don't assume

An eg65 is the right boiler. Measure all the radiators to size the new boiler when that day comes.

Binnacle

of course

Yes, I do know that.  But like I said I'm not buying so for quick comparison I reference the closest thing to what I have.  Based on the way the system behaves it's probably correct, but I will do the calculation if replacement becomes necessary.



Reading a lot about this and it seems that unless some water-related pathology exists there is no easy way to calculate life expectancy for well made boilers.  It could go another 20 years--or fail tomorrow, but 10-to-20 years is more likely.

Binnacle

oil in the boiler, for sure

That union in the relief valve tapping has been terrifically useful.



After a couple of days I opened it up and lo!  a Thin Film of Black Oil has coated the inside of the pressure valve.  So much oil that two or three drops oozed from inside onto the rim of the union surface.



So for 50 freakin' years the boiler has needed to be skimmed.  Probably a rush install as they used 2 inch pipe and a bushing off the top of the boiler even though the install manual specifies 3 inch pipe (no steam header required, explicit in the document).  And my suspicion that the skim port was never touched is clearly founded.



I'm guessing the previous owners (one family the entire 87 years) never hired a competent service professional or decided not to spend the money when skimming was recommended.



I suppose it's a testament to the sturdiness of steam technology that systems work tolerably well and with decent efficiency even with a slew of unresolved problems.



So I did a quickie skim using some clear plastic tubing, pulling about five gallons off the top.  Then added a third and fourth Steamaster tablet, ran it for one cycle, blew it down from the bottom drain to about half full, let it cool for three hours and finished draining it.  Washed a fair amount of junk out of the near end of the wet return as well.  Will have to flush that this summer.



Then filled it from the relief port with hot water (to minimize thermal stress), steamed it one cycle to get rid of O2 and two Steamaster tablets for the 19 gallons of water.



Now the glass is essentially clear, with just a hint of cloudiness and that pretty violet hue.  Seems like the Steamaster tablets cause most free rust particles to go into solution as there are now none visible in the glass and there was hardly any rust in the blowdown and drain water this time (wet return drain is separate and was ugly).  Less water running down the inside of the glass at the beginning of each run and once the level gauge fittings get hot and stop condensing steam, none at all.



This spring I'll take out that plug, put in a nipple+cap and give the boiler a proper skimming.

ALIGA

do you have pics of the near boiler piping?

might be a good task to redo in summer as well with a drop header.

Binnacle

pipes

Sure.



I don't think I'm likely to re-pipe it.  Seems quite a large project and several other important items such as proper main and riser insulation and flushing the wet-return are less difficult and provide bigger return on effort.



I've seen posts where this type of classic older boiler design is described has having a "deep chest".  IE substantial clearance above the water line for steam to form.  Is said that only newer high-efficiency boilers are highly sensitive to the header assembly (to the point where an installer would be in trouble with the manufacture for non-compliance) since the boiler chamber size is minimized.



As you can see the boiler cover was recently removed and reinstalled, probably because of a leak.  Because of this I feel the risk of trouble in removing the skim port plug is acceptable as the bolts are ok and the cover can be removed and the gasket or cement redone if the seal breaks.  Unfortunately the rear pipe (Hartford loop?) is 3-3/8 from the cover plate and insufficient clearance is present for installing a street elbow or tee, so I'm stuck with installing a short nipple and cap (really don't want to touch those existing pipes) and using a large plastic hose for skimming.



Also I probably should bring in a professional who has an analyzer to tune the burner intakes.  The rush-job installer left them set wide open.



regarding the photos:



* the bulge in the pipe wrap is union

* the foreground pipe in the side-back view is the gas line

* note the new bolts on the boiler cover

* string off gauge tree is a hack to provide tension to minimize wobble



boiler manual:



http://www.heatinghelp.com/files/posts/21340/Weil-McLain_E-Series-1-Gas.pdf

BobC

I've seen worse

My old Burhams header was lower than that and worked fine, having the header higher and bigger would be nice but I don't think I'd mess with it unless the problems were severe.



When the time comes to replace the boiler I'd do it up right.



Bob

Binnacle

-

Indeed, the WM manual says it should be 3 inch pipe at least 2 feet high.



But it all works well enough and I'm not a masochist.  I like all the little fixes an amateur can easily make and result in big improvements.  Getting the boiler clean and generating drier steam seems enough.  Fun stuff that's not painful or treacherous.



No doubt I'll have it done right if ever it needs replacing.  My house inspector suggested two units, one for each floor/apt but with a gas bill that maxes at $200/month I don't see how it's worth the expense to run a second steam main and rework half the radiator run-outs or switching to hot water and replacing everything.

Binnacle

MOVING AHEAD -- science is my guide

I've been skeptical about the idea of TSP cleaning being much of a risk, even for an old boiler.



So I set out to apply some science to the matter, and have now concluded that the risk of a problem, even with an 50+ year old WM steam boiler is remote.



What stood out was that I could not find one single horror story after extensive web searching.  (This is the cue for anyone with TSP horror stories to come forward.)



Reading about boilers, I learned that an incorrect acid/base water balance is one of the biggest (but not the only) danger to the integrity of the metal. Boiler water that is too acidic (lower pH) promotes iron to oxidize rapidly (i.e rust). With overly basic (high pH) water, the excessive alkalinity has a caustic corroding effect. Unlike human skin (we don't soak our hands in lye, or even for long in soap), iron fairs best with a substantially alkaline water balance. At a pH of 10.2 iron does not corrode at all and exists in equilibrium with vessel water. A slightly higher pH of 10.5 through 11 has an added benefit of combating other water-related issues to the advantage of the system. When the water pH reaches a level of 12 for an extended period the boiler metal can become brittle and form microscopic (or not so microscopic) cracks. Higher than 12 and caustic dissolves and gouges the interior of boiler.



First I obtained an alkalinity test kit:



http://www.amazon.com/Hanna-Instruments-HI3811-Alkalinity-Approximate/dp/B0085WWNJ6



Then some high-quality pH test paper:



https://www.microessentiallab.com/ProductInfo/F01-SHTRG-090120-SRD.aspx

https://www.microessentiallab.com/ProductInfo/F04-MCRFN-092106-MID.aspx

https://www.microessentiallab.com/ProductInfo/F04-MCRFN-102123-MID.aspx

-----got the one below, but should have purchased the above one

https://www.microessentiallab.com/ProductInfo/F01-SHTRG-102120-SRD.aspx



from LabSource.com



60784317 TEST PAPER PH 9-12

60775972 PH TEST PAPER 9.2-10.6

60784361 TEST PAPER PH 10.2-12.0



Then I discovered that with just the three Steammaster tablets in the 19 gallon boiler the OH (hydroxide) alkalinity was about 95 ppm. Dropped to 70 ppm on a small blowdown and make-up of 0.5 gallons. This is below the recommend 150-300 ppm OH alkalinity for low-pressure boilers. Total alkalinity was 150 ppm which is low. The first-try cheap pH paper from China put the pH between 9 and 10--too low.



So I added three teaspoons of pure sodium hydroxide (thinking the calculated one teaspoon could not possibly be enough), and wow, the caustic OH alkalinity shot up to 400 ppm. So next time it's one teaspoon. For now I blew down and made-up one gallon and the OH level dropped to 300 ppm. Seems ok to let it ride awhile at the high side of the range since this has a cleaning effect. pH is pretty close to 11.0 which is generally considered ideal for steam boilers.



Having spent a few days playing with boiler chemistry, went back to the TSP question. Careful reading of the Weil McLain document shows that they are actually recommending ONE cup for a 200 MBH boiler (in their words "1 lb per 50 gallons"), not the oft-repeated one cup per 100 MBH.



That's a 0.25% TSP solution rather than a 0.50% TSP solution. The Wikipedia article on TSP says that a 1% solution has a pH of 12. High for a boiler but not crazy high--how bad could a 0.25% solution be?



Adding 0.8 teaspoons of TSP to a quart of water (per WM's scaled-down advice) yields water that has a pH of a bit over 10.5, probably is 10.7. Not scary at all. What's more, the OH alkalinity of this solution is 180 ppm--comfortably at the low end of the 150-300 ppm OH ideal range. Total alkalinity (per the bromophenol titration) is 625 ppm, which is less than the maximum recommended operating alkalinity of 700 ppm.



My conclusion is that running a 0.25% TSP cleaning for two hours at 180 degrees per the Weil McLain advice poses no more risk to the integrity of the boiler than operating it normally with proper OH and total alkalinity.



Also performed accurate testing on the tap water and found it biases slightly alkaline. pH is about 7.5 and the bromophenol test shows 55 ppm total alkalinity (phenolpthalein alkalinity and therefore OH alkalinity are zero, as of course they must be below a pH of 8.3). Per my observation, boiler water here tends to become increasingly alkaline as water is lost through evaporation. These alkaline biases (plus WM's old-school over-engineered 800lbs of cast iron) are probably what has kept the beast alive and well for 50 years.



It's predicted to be 50 degrees for the next two or three days and so I'm going to run a TSP cleaning and get the stupid oil out of the boiler at last!



Will of course start with fresh neutral water and check the chemistry. Even have a lab thermometer to lower into the relief-valve union for checking the temperature.



NOTES:



1) TSP used is Savogran brand found at Home Depot. Is actually 75-80% trisodium phosphate and 20-25% sodium sesquicarbonate. The latter component is another common caustic additive and likely accounts for the 180 ppm of OH alkalinity.



2) Boiler chemistry recommendation taken from the Alberta government operational guidelines for commercial boilers and checked against additional sources. Please no one suggest that a residential boiler is somehow a different animal than a commercial boiler. The only difference is that residential boilers are smaller and are almost always neglected where commercial boilers are bigger and (sometimes) operated by professionals. Document attached.



3) OH = 2*P-M



OH = calculated hydroxide alkalinity

P = phenolpthalein titration alkalinity

M = bromophenol titration alkalinity



4) with water 1 ppm equals, roughly, 1 milligram/liter



5) Steamaster tables are primarily sodium nitrite, which is a rust inhibitor and a mild alkalizing agent



6) The phosphate in TSP reacts with oil and residual minerals from tap water and facilitates their removal when the boiler is drained; at one time was the key ingredient in laundry detergent



7) Alberta recommends keeping 60 ppm of phosphate in boiler water. So I calculate I'll add one teaspoon of TSP to the new boiler fill-water after cleaning. Will add that with the Steamaster tables before testing the alkalinity and bringing OH into spec with sodium hydroxide.

Binnacle

0.25% is safe, 0.50% perhaps not so much

Decided on some further preparatory experimentation:



Placed one quart of water in a badly-burned and corroded, heavy-walled sauce pan and heated it to between 180 and 190 degrees. Added four drops of SAE 20 non-detergent oil to the surface.



The oil spread out in a thin rainbow colored sheen.



Then added one teaspoon of TSP. When the TSP hit the water it reacted and turned white for about fifteen seconds, then settled down to a calm clear state.



The oil on the surface broke up and it appeared that much of it dissolved. A few small patches remained floating about, indicating that ideally one still should skim as part of a TSP cleaning when possible.



Checked the alkalinity and pH and arrived at the same values as with the hot tap water experiment earlier.



This is the 0.25% WM recommend TSP cleaning setup.



Now to make things interesting, I added a second teaspoon of TSP to bring the concentration to 0.50%. This is equivalent to the frequently recommended one cup per 100 MBH (10 gallons of water) cleaning procedure.



Had little effect on the surface oil, perhaps a bit more went into solution.



However the pH shot up to perhaps 12, possibly higher. More impressive, both the P and M alkalinity went ballistic. P-alkalinity went to 1000 ppm and M-alkalinity went beyond 2000 ppm--I gave up on the titration to save solution.



This demonstrates as perfectly valid the cautious advice provided earlier in this posting. I'm comfortable with 4oz of TSP per 100 MBH, but 8oz pushes the water into high causticity and I'm not going to unleash that on the aged 800 lb iron pot in the basement.



Anyone considering a TSP cleaning on an older system would be wise to spend $30 on an alkalinity test and check what they are doing. I might add an extra ounce or so of TSP to bring the pH up closer to 11.0, but the causticity curve looks to be steep and tricky.



My thinking is it would be safer to perform two or even three 0.25% TSP cleanings on an older system than to hit it with one 0.50% cleaning.



After awhile a cloud of white precipitate accumulated at the bottom of the pan. Suspect this is neutralized undesirable substances now bound up with the phosphate, but I don't know for certain.



And that burned saucepan is quite improved, though sadly not restored.

Binnacle

TSP used properly is super-safe, even for old boilers

Starting with the results: Brilliant! Steam is much drier and the entire system sounds dramatically better. Sound of water boiling in the vessel is mellow and almost inaudible. Much less swooshing in the risers and radiators. Sound of steam entering the most troublesome radiator actually sounds dry, if such a thing is possible. Bouncing of level-gauge water is almost gone. Probably the planned spring opening of the skim port and skimming will move the boiler's behavior into the realm of perfection.



Now for the effort:



Let the boiler cool for three hours after a cycle, water temp was 130 degrees, then drained and filled it with fresh hot water via the top relief-valve union. Not much rust and other junk in the water due to earlier drains and fills and keeping the pH between 9 and 11 the last few weeks.



Added five ounces TSP and ended up with a pH of 10, P-alkalinity of 222, M-alkalinity of 660 leaving OH-alkalinity zero with a big P*2-M negative bias of 216.



So it occurred to me that the two-cup prescription takes into account that much of the TSP will immediately react with oil and other contaminants and the high alkalinity seen with pure water and 0.25% TSP in a pot is not representative. Decided to go big and ended up adding 20 ounces (2.5 cups) of TSP in two increments. Ended up with



P 672

M 1665

OH 0

pH 11



which still has a substantial way to go before becoming caustic. Found a recommendation for cleaning with 5000 ppm TSP and 20oz in 20

gallons is 6000 ppm--not way off that



http://www.steamesteem.com/feedwater/boiling-out.html



What makes TSP safe for an old system is a) not supplementing TSP with caustic lye (sodium hydroxide) and b) monitoring the pH and not letting it get past 11 or so. What's happening here is that the boiler is, in effect, being washed out with mild detergent solution.



Kept the water at between 190 and 200 degree for two hours after the final TSP addition, and the entire cleaning ran for about four hours.



When draining saw virtually zero particles and no rust whatsoever--almost perfectly clean. Drain water had a dark tinge indicating, I believe, dissolved oil.



Discovered through recent reading that TSP is considered an excellent boiler water additive. Strong positive experience reported right here on HH:



http://www.heatinghelp.com/forum-thread/91982/Sodium-Nitrate-in-Steam-Boilers



The US Navy uses primarily TSP and some additional soda ash for all marine boilers, preferring to keep the pH close to 10.5 to minimize the risk of tiny eddies of caustic water in the nooks and crannies of their large boilers. I prefer a pH of 11 for the small low-pressure boiler here as do many others. Keeping the concentration of phosphate in the boiler at around 60 ppm does an excellent job of protecting the boiler from various forms of water-related nastiness.



After filling I added three ounces of Savogran to obtain about 60 ppm of phosphate (6.9g per 4oz noted on side of box) plus three Steamaster tables. After a day two teaspoons sodium hydroxide dissolved in a quart of water to bring the caustic up. Now have



P 342

M 600

OH 84

pH 11



and after another day plan on another teaspoon of sodium hydroxide or two to bring caustic into the 150-300 ppm range.



I've decide that sodium nitrate from the Steamaster tables is superfluous here since I will monitor the pH and anything over 9 or so completely inhibits oxidation regardless of how the pH is achieved. Also switching to pure TSP with the idea that total alkalinity can be kept lower with optimal phosphate and hydroxide alkalinity if the sodium nitrate and sodium sesquicarbonate are eliminated. So in a week or so I'll drain and refill again, adding just pure TSP (on order) and sodium hydroxide. Draining after running with some phosphate should knock down further any oil residual left behind by the cleaning. As I understand it the oil will end up as soap in solution and depart peaceably with the drainwater.



I would still highly recommend Steamaster tablets as the treatment of choice for the less-obsessed. The sodium nitrite works by scavenging dissolved oxygen in boiler water, thus inhibiting rust, and by biasing the pH to somewhere between 9 and 10, a reasonable range especially for the typical situation where no one is monitoring the water. However I agree with the HH consensus that one-to-two tablets per 100 MHB or 10 gallons of water is the correct dose and the Steamaster label directions are excessive.



A note for anyone who contemplates a similar approach: Even one teaspoon of sodium hydroxide dissolved in a quart of water results in a highly caustic solution with a pH up around 14. Even when pored into vessel and mixed with the existing boiler water, the concentration might still be high enough to do some damage. So one should always follow the addition of hydroxide solution with a fair amount of make-up water or recycled blow-down water to disperse the caustic solution quickly. Then immediately cycle the boiler to further disperse the hydroxide. The Navy considers it too risky to use sodium hydroxide for the aforementioned reasons. This makes sense considering the high capital cost of their equipment and the extreme rate of job assignment turnover in ranks as crew are trained for various tasks and rotated constantly. Too easy for inexperienced or careless crew members to damage a boiler.



Attached are a graph that show the fundamental pH chemistry of water (found at http://www.sprungtraining.com/messageboards/tabid/73/forumid/14/postid/175/view/topic/default.aspx) and Navy document showing the effects of pH in boilers. In addition here is a helpful paragraph from a Navy boiler operations manual. Much additional information may be found by searching on "COPHOS boiler".



https://www.google.com/search?q=COPHOS+boiler



-----



Boiler Water



The boiler acts as a concentrator for all the materials which the feed system pours into it. All contaminants remain in the boiler and produce damaging conditions. In order to minimize these, boiler water treatment programs have been established. The present standard Navy treatment is based on a regimen known as coordinate phosphate – pH control (more simply, COPHOS) to serve several purposes:



1. Maintain the pH and phosphate levels so that caustic corrosion cannot occur.



2. Maintain the pH sufficiently high to limit corrosion and to protect against acid-forming magnesium reactions by forming magnesium hydroxide (Mg(OH)2) sludge.



3. Maintain a phosphate residual in the water sufficient to precipitate calcium and magnesium as phosphate sludges which are less adherent than scale.

Canucker

Experiment

I like your research into this, been very interesting. The precipitates forming in your saucepan after the NaOH was added are most likely the chlorides that form from raising or lowering your solutions pH. And you and the Navy are correct in not wanting to use too high pH solution(or lower for that matter) and using water to adjust because it takes a factor of 10 for each level you wish to raise or lower the solution i.e; 1 litre of pH 4 solution to pH 6 requires approx 100 L of water, to hit 7 requires 1000 L. It's one of the reasons that we don't use water to adjust pH if there is a spill at our facility. Massive increase in size of the clean up

Pumpguy

Chemistry

I found this thread interesting although the chemistry detail is a bit over my head. 



My involvement with steam heating systems is almost always on 2 pipe systems with condensate and vacuum pumps applied. 



Many times I have seen the bronze pump parts badly corroded due to the improper use of boiler water treatment chemicals, and where they are introduced into the system.  On these same systems, the cast iron and carbon steel was in good condition.



My sources tell me this corrosion is due to the high alkalinity of the boiler water.  I have been told the pH should be alkaline, but no higher than around 8.5. 



After reading this thread, it would appear that there is much more involved than merely keeping a certain pH level.



Can anyone comment  on how boiler water treatment chemicals affect the bronze pump parts, along with the cast iron and steel of the rest of the system?

Binnacle

isolated?

Are the bronze pumps you reference for bringing return water back to the boiler?  Return water is distilled water and would be slightly acidic.  Without knowing much about two-pipe return setups I would imagine the return water would be at least somewhat isolated from the boiler vessel so and the alkalinity of the boiler water would not affect the return lines much.  Perhaps a separate return-line treatment regimen is called for.  Much of the high-end boiler operation literature I've seen describes various types of return- and feed- water treatments.



The single-pipe system gravity wet-return here connects to the boiler piping loop at just below the water surface level, so only slight diffusion of boiler water into the returns occurs.



Assumption here is you mean a two-pipe steam system.  Hot-water systems are different creatures and water treatment much different.

Pumpguy

Yes, 2 pipe systems.

On 2 pipe systems with pumped boiler feed, the condensate flows by gravity to to the condensate return or boiler feed pump where it is pumped to the boiler(s).  Occasionally a vacuum pump is used to lift condensate up from lower return lines.  The vacuum pump also removes air and lowers the pressure in the return lines. 



I have also many times seen where the condensate has been allowed to go acidic, and that damages the cast iron and steel.  I have seen cast iron so damaged from acid that you could take a small tool and dig deep gouges into it.  The iron turned into a carbon / graphite paste. 



So, obviously condensate chemistry is important.  This thread addresses how the chemistry affects cast iron and carbon steel.  Since condensate, boiler feed, and vacuum pumps are almost universally  bronze fitted, I would like to learn more about proper chemical treatment to prevent corrosion problems, and what advice to give my customers.

Jean-David Beyer

takes a factor of 10 for each level you wish to raise or lower the solution

My boiler has an aluminum heat exchanger (hot water, not steam) so I cannot use TSP in it.



However adjusting pH can be tricky, depending on what is in the water already.

If the boiler water were too acid, one might put in some sodium hydroxide (preferably already in a solution of known concentration. Now if the water in the boiler was just plain water, and you wanted to raise the pH of the boiler, the amount of sodium hydroxide could be calculated as you described.



And if you went too high, you could titrate it back with hydrochloric acid.



But there are solutions, known as buffer solutions, where they resist (within limits) any significant change in pH.



For example, a solution of disodium phosphate and sodium hydroxide wants to hold a pH of about 11.5/

A solution of sodium bicarbonate and sodium carbonate wants to hold a pH of about 10.2

A solution of boric acid and sodium hydroxide wants to hold a pH of 9.2.



Now if you put lots of acid into a solution, the pH will eventually go down as the existing mixture is changed.



I would imagine those who sell chemistry for adjusting boiler pH actually supply buffer solutions so that the pH does not change for a longer period of time that a simpler solution would.

Binnacle

variety of options, cost and practicality main issue

Here's an interesting link to a GE division's web-site discussing various pre-boiler corrosion control strategies, though I believe the target audience is commercial continuous flow boiler operators:



http://www.gewater.com/handbook/boiler_water_systems/ch_11_preboiler.jsp



I don't know enough about any of it to suggest a particular approach except to say I've seen frequent references to raising the temperature of return-line water with a heat-exchanger or independent heat source to drive off excess oxygen. With bronze and brass that might be one of the better approaches since it appears that having a near-neutral pH is best for these metals. [disclaimer: just quick-searched that and it could be incorrect]



An amazing amount of information is available to the persistent web-searcher, but boiler technology is not mainstream so one has to work a bit when composing searches. For example, the above link to searching for "COPHOS boiler" will pull up a great deal of US Navy boiler documentation and even research because COPHOS is a highly specific Navy acronym used only by that institution. This has particular value because the Navy has been operating a large number of boilers for a long time and qualifies as a higher-authority on the subject. So I would point you in the direction of spending a few evenings whacking away with Google at the topic. This is what I did when trying to figure out the story with TSP cleaning, ending up with the bonus information regarding TSP as probably the best steam boiler additive available.



Best of luck on this. Perhaps others with more specific knowledge and experience than I have will chime in on this thread and offer further assistance as well.

Binnacle

alkalinity

What you describe is referred to as alkalinity.



The phenolphthalein and bromophenol blue titrations referenced above measure this. A quick web-search will return much on the subject. The $30 Hanna kit linked does an adequate job for boiler operation, though greater precision can be obtained with proper laboratory glassware. A slightly more involved two-step titration is available for directly measuring hydroxide alkalinity.

Binnacle

Erratum

While calculating the quantity of pure TSP (as opposed to the Savogran blend) required to obtain the desired 80 ppm concentration of phosphorus, I discovered a glaring error in the original calculation.



Pure TSP is 1/12th (8%) phosphorus by weight.  Savogran is 6.1% elemental phosphorus as declared on the side of the box and this jives with the material data-sheet description.  The box also states that each ounce contains 6.9 grams of phosphorus, but they mean ounces as a unit of weight, not as a unit of volume and I took the statement to indicate fluid ounces.



The density of TSP-dodecahydrate is 1.620 g/ml, so one ounce by weight is 0.5917 fluid ounces, the kind of ounces marked on the side of a measuring cup.  [Note the TSP+sesquicarbonate blend has a slightly different density.]



So I added 50% or so more than the intended quantity of TSP and have 120 ppm of phosphorus floating around in the boiler.  Not so terrible as this will likely scrub the last traces of oil out of the system and the plan is to refill and add only TSP and sodium hydroxide as described earlier.



For anyone interested in the math, the details below.  Relied heavily on the Google search box to perform unit conversions such as milliliters to fluid ounces and Wikipedia for the chemical attributes of TSP.



TSP additive, 80ppm phosphorus-O4 in 20 gallons

  TSP molar mass 163.94 g/mol

  12(H2O) [dodecahydrate] molar mass 12 * 18.01528 = 216.18336 g/mol

  total 380.1233 g/mol

  phosphorous molar mass is 30.9737622 g/mol

  TSP dodecahydrate is 30.9737622 / 380.1233

    or 1/12.272431 phosphorous

    by weight or mass

  want 80 mg/l phosphorous in 75.7082 liters (20 gallons)

    or 6.056656 grams phosphorous

    or 6.056656 * 12.272431 = 74.329887 grams TSP-d

    or 74.329887 g / 1.620 g/cc = 45.88265 cc TSP-d

    or 1.55148 fluid ounces

Binnacle

perhaps ask the equipment manufacturers

A quick look turned up



http://bellgossett.com/literature/



where several documents describe some of B&G's various condensate handling components.



The B&G stuff looks rather high-end, but it gave me the idea to suggest asking the manufactures of the equipment you work with for advice and suggestions.

Binnacle

Feed Water pH

Came across this while looking for something else:



Feed Water pH

• Feed Water pH should be above 8.5 to prevent corrosion of iron within the system.

• Feed Water pH should be below 9.2 to prevent copper loss within the system.

• Optimum Feed Water pH should be 8.5-9.2.



page 31 the following PDF document



II. BASIC BOILER WATER PROBLEMS, OXYGEN CONTROL EQUIPMENT and CHEMISTRIES



http://www.kurimexicana.com/pdf/AWT-2.pdf

Binnacle

TSP rocks. . .

. . .as a cleaner and as an additive.



Final results are amazing: Crystal clear water and only small quantities of black magnetite (the good iron oxide Fe3O4) particles in initial LWCO blowdown water followed by clear clean water. Not a trace of hematite (the not-so-good red-rust iron oxide Fe2O3). Zero residue in the pristine white blowdown bucket. Before-and-after pictures of the level gauge attached below.



Drained and refilled again after the first post-TSP-cleaning fill described above.



Added 60 cc's of TSP and one teaspoon of sodium hydroxide for 20 gallons, 200 MBH and ended up with



P alkalinity 273 ppm

M alkalinity 372 ppm

OH alkalinity 174 ppm

M-P alkalinity 95 ppm



Forty-six cc's of TSP in distilled water would produce the desired 80 ppm of phosphate, but an additional fifteen cc's were required to react-out tap-water hardness.



Three different pH papers read 11.2, 11.0 & 11.7 but this is only approximate. Different papers seem to respond to different sources of alkalinity (the paper showing 11.7 appears oversensitive to hydroxide alkalinity). Exact pH is not critical when alkalinity is monitored so I'm not planning to purchase one of the rather expensive pH meters. Before the sodium hydroxide was added pH gradually rose from low 10's to the mid-to-high 10 range over a four day period.



In addition it's obvious that the inside of the boiler is developing a surface layer of magnetite. Magnetite is strongly protective against further corrosion and intentionally establishing this layer is the purpose of maintaining hydroxide alkalinity in the range of 150-300 ppm. Can see the magnetite layer forming on the inside of the pressure relief valve pipe and on the union surfaces.



Earlier I mistook magnetite particles in solution for oil. Gives water a dark tinge. However I'm sure oil was in the boiler before the TSP cleaning because the black drops observed felt like oil. Was oil with magnetite particles. One must rely on texture and/or observation of floating surface blobs to distinguish oil.



Alkalinity will increase as water evaporates so won't be adding more than the one teaspoon of sodium hydroxide for now. The initial TSP charge established OH alkalinity of roughly 100 ppm. My approximate understanding is the phosphate reacts with and neutralizes magnesium and calcium and the result is precipitant sludge and hydroxide. So as local water rates harder and harder is, more and more TSP will be required to purge the water and establish the 40-80 ppm of recommended phosphate alkalinity. My observation is that the M-P difference value appears to match the level of phosphate. This is approximate as carbonate is also included by this calculation, but since only pure TSP was added, almost all of the M-P number is representative of phosphate. (Found this in a Navy document described as an "emergency phosphate test."  At some point I suspect I'll purchase the Taylor K-1109 phosphate test (http://www.novatech-usa.com/K-1109) though it's a bit expensive.



Have observed that it can take as long as three days for additives to be fully represented in alkalinity tests of water sampled from the level-gauge pet-cock. So one must be especially careful when adding sodium hydroxide, err on the side of less rather than more, and be patient.



One can add approximately one fluid ounce of TSP and no sodium hydroxide per 10 gallons of boiler water when filling and just forget it, without bothering to ever test the water. Several comments to this effect appear on HH. The Navy states that it is nearly impossible to end up with overly caustic water when using TSP alone. My observation is that a low, safe level of OH appears with TSP which provides the antioxidative benefit of alkalinity and promotes the formation of a protective magnetite layer.



My next adventure in water chemistry is determining whether sodium sulfite can be used for the low-end residential boiler here as an oxygen scavenging agent. Have read that unless incoming return water is quite low in O2 so much sodium sulfite will have to be added that it pushes the TDS (total dissolved solids) unacceptably high. With commercial boilers feedwater is usually processed by a deaerator of some sort, but with a small boiler the only thing driving off O2 in the return is the warm-to-hot temperature of freshly arrived condensate and the limited exposure to air offered by the wet-return pipe.



If sodium sulfite works it will be great as this should bring oxidation of the boiler surfaces down to virtually zero. If not then it's drain-and-fill and either be satisfied with the benefits of TSP+NaOH alone (which worked for decades for the Navy) or go exotic and look into EDTA, DEHA or perhaps even hydrazine (scary stuff though).



Starting TDS is 885 ppm, leaving quite a bit of headroom to the Alberta recommended maximum of 3000 ppm. The Alberta document implies that with sulfite treatement a minimum of 1500 ppm should be expected. Added three teaspoons (one teaspoon should produce 60 ppm in the total absence of O2). Tomorrow will tell. . .

Pumpguy

Feedwater pH

Thanks Binnacle!  That's useful information I need.

Pumpguy

Had a chance

to go through that website and found it to be the most informative I have seen on the subject of boiler water treatment chemicals and corrosion protection.



Back in my service technician days I visited many large fossil fueled power plants that had big Nash vacuum pumps applied to the steam side of the main condenser.  These vacuum pumps had to be all iron.  If any copper bearing parts were fitted, they would corrode quickly away.  I was told this was due to the Hydrazine in the water. 



Now I know what Hydrazine is, and why it is used.  Thanks again for posting this information.  It is very much appreciated. 

Binnacle

sulfite a bust; finished with water chemistry

I wasn't too surprised to find that adding sodium sulfite as an oxygen scavenger is as other have pointed out "futile".  The boiler ate 100 ppm a day of sulfite an the resulting sulfate pushed TDS up at, well about 100 ppm a day.



So that doesn't fly.  Then I thought about the fancier oxygen scavengers and realized they are all volatile compounds that travel with steam up into the system.  They range from highly carcinogenic (hydrazine) to safe for industrial boilers but still not considered safe for food-applications (DEHA, EDTA, and other amines).



Therefore I'm just fine with ongoing corrosion in the steam pipes and return lines, and the smidgen of corrosion in the boiler that the phosphate+hydroxide treatment cannot perfectly arrest.



I may fill the boiler to the top, add a teaspoon (100 ppm) of sodium hydroxide, cook out the O2 and dose it with 250 ppm of sodium sulfite at the end of the heating season.  This is a commonly applied corrosion inhibiting "wet layup".



-----



Drained and refilled one last time to put TDS back where it started at 900 ppm and it looks like this also exorcised the last traces of saponified oil.  Steam is a bit drier and the pressure gauge and level gauge are tamed to no more than a minor quiver at the start of cycles.



Attached a picture of the initial blowdown after the refill.  Shows soft, non-adhering phosphate-bound sludge reacted out from the fill water.  Has a bit of magnetite coloration.

New Hampshire DIYer

How much Savogran TSP?

Thanks so much for all your thorough work. It's very helpful and appreciated. I recently had to put in some Hercules boiler stop-leak which, in addition to fixing the leak, appeared to clean the boiler. The water in the sight glass looked like the stuff you find at the bottom of a red wine bottle--dark reddish brown and full of sediment. Anyway, now that I've drained and re-filled the boiler, I'd like to boost pH to prevent corrosion.



If I read your posts correctly, I could use the Savogran (that's already in my basement) like this:

Assuming the density of the Savogran is close to the 1 oz = .5917 fl oz you stated for pure TSP, I would need a total of 3.11 oz of Savogran. Here's my math:



1/.5917 = 1.69 dry ounces per 1 fluid oz of pure TSP

Divide 1.69 oz by .76 or the percentage of TSP in Savogran (6.1% vs 8% in pure TSP)

= 2.22 dry oz

x 1.4 for 14 gallons of water in my boiler

=3.11 oz.



Any problems with this approach?

Perhaps I should also mention that I'm on a city water system that sends water out from the treatment plant with pH adjusted to between 9 and 9.5.

Binnacle

maybe a little too much

Without digging into the math presented, I'd say it's a little high unless you want to do a "cleaning fill" where you intend to run it for a week or two with extra TSP and then drain & fill it again.



In the 20 gallon unit here I used 3 fluid ounces of Savogran on the first try, but it worked out to 120 ppm which is a bit higher than the recommended upper limit of 80 ppm of phosphate.  I used this as a cleaning fill that was replaced subsequently.  Note some of the TSP is consumed immediately in neutralizing magnesium and calcium so probably 1/3 of an ounce is eliminated there.



For longer term I'd recommend about 2 fluid ounces Savogran for 14 gallons.  This still includes a surplus for initial water neutralization and should get you about 80 ppm and a pH in the mid-to-high 10's.

New Hampshire DIYer

Confused

Maybe I'm having a hard time because of the confusion between dry and fluid ounces. I'm just trying to figure out what weight of Savogran to put in the boiler as a long-term water conditioner. If you're suggesting two fluid ounces, doesn't that equate to roughly 3.38 dry ounces of Savogran? (2 x 1.69)