ML20069F620

Testimony of a Savage on Contention 4 Re Corrosion & Sludging of Condenser Tubes by Pond Water.Prof Qualifications Encl

ML20069F620
Person / Time
Site:	Midland
Issue date:	03/18/1983
From:	Savage A SINCLAIR, M.P.
To:
Shared Package
ML20069F603	List: ML20069F601 ML20069F620
References
ISSUANCES-OL, ISSUANCES-OM, NUDOCS 8303230248
Download: ML20069F620 (8)

Text

- . . .-

i Statement of A.B. Savage as to Sinclair Contentionumber N 4i Conclusions i

The corrosion and sludging of condenser tubes by pond leakage of pond water of variable composition into water themayw t induce secondary cooling system at a driving force of perhaps a er of the 50 psig A much more serious cause of steam generator corrosion that unit which supplies steam to The Dow Chemical , particularly of.

ompany, will be the chemical industrycareless operation of the cation and anion

, very common in exch the use of ammonia for neut lization and.of hydratine for oxygen s,cavanging, and failure to establish adequate controls for when blowdown should be performed.

Contact of dissimilar metals of the tube and shell comp steam generator with the electrolyte will set uponents .

of the coupl corrosive es.

initiation of intergran'tlar corrosion. Theare chemical aid in the well suited to the materials of construction. not liquid and vapor pressures during impingement at high velocitiesS operation vibration, tress failure at a driving force ,of 1,000 will contribute psig. to steam generator tube, transie Such tube failure will necessitate. shutdown. The of the shut-unit Considering the result could have been a disaster.

the narrow margin of safety and as ressure, erratic. to p The time.

with incidence of tube failures in steam generators isncreasing i Improved materials of construction improved design and fabrication any more units to start up.and improved water chemistry areorne,ed 4

T

~ '

8 ,

l l

The sketch is inadequate and confusing'. Reference . Hillman. is ma

1. Frimary coolant.

The primary through coolant the tubes is shown of the boiler circulating up through the reactor , down the reactor, at a stated pressur,e of 2,000 psigand through a centrifugal pu accumulate above the upper tube sheet in the generator.. Any gases will tend to

2. Steam flow.

Feed water is introduced into and preheated in the lower generator, vaporized in,the shell, and the steam leaves annulus the uppe of the anhulus ofin'the generators generator below the top. The steam flows torturbo tandem -

vacuum conditions, an,d is recycledwhere it expanded and ultimately condensed by a centrifugal pump through a

" full-flow" demineralizer and back into the annulus ofenerator.

the g

3. Condensation.

The vapor leaving the' turbines is condensed in the shell .

of the cond-water) from the pond passes through the tubes ontal condenser, and is returned to the pond. y horiz-of ce the

4. Steam sold to The Dow Chemical Company.

In NUREG 0537, steam production of from 1.4x10 6 steam per hour is stated, or from 46 to 4.05x106 ounds of perhaps an average <or 109 gal./sec.. 5 gal./sec. to 134 gal. sec., or reactor #2 will produce 825 MV, difference, of feed water. It is stated that 148 MV of power equivalent will go to steam: reactor #1 504 MV o 3.6x10 lb./hr. at 175 psig., plus 0.4x10 lb./hr. at 600 psig.

The point where the steam is withdrawn is not stated .

turbogenerator stages:One may suppose that the steam is withdrawn int e ween two 600 psig.(approximately 485 F throughmately .

a turbines 575 F) to and(

(b) Withdrawal of 600 psig. steam for Dow at this point.

('c) Stage 2. Expansion of sgeam at 600 psig.(approximat ey 485 l 175 psig.(approximately 375 F) through a turbine; F) to (d) Withdrawal of 175 psig. steam for Dow at this point

( e) Stage 3. Expansion of steam at 175 psig. (approximat' l .

a turbine to 20 inches of mercury vacuum ,1900F. e y 375 F) through (f) Condensation of the remaining steam at this pressu re.

(g) The vacuum powered vacuumunit,could be maintained by an Ingersoll n Ra d t from the system, ype steam jet discharging through the. hot well, and removing air

5. Makeup.

Perhaps 80 per cent of the steam sold to Dow would be ret amounting to a maximum810,000 ofensate and reused. The balance urned as cond-be mad would This amount of water would be passed through ion eexchangpounds .

purification per units. It would also be used for primary or the secondary cooling system. makeup to offset any leakage om t.he fr

3

6. Water cuality. .

(a) Pond, or service water.

The water pumped from the pond will contain algae , salts ent, contaminants from The Dow Chemical through a set of bars and through a magnetic separato en passed a onComp

, sedime

r. The listed side of the condenser and to contribute to its corr e inlet on.As the pond water will come from the Tittibawassee Reiver, it will c 30 ppm. of Nacl. One of the upstream tributaries, theer ontain Salt perhaps Riv contain as much as 90 ppm. of salt.0therwise the river can sandy, low mineral area. It may contain organ,ic materials comes from fr a floor, salt from highways, agricultural contaminant om the fores' the pond is not s'tated :so concentrations are s,etc.t iThe volume of unce vary widely. Concentrations will depend upon the volume r a n. Doubtless they ative locations of inlet and upon the tel-blowdown discharged. The pond will contain acid andwithdrawal site by the exchangers. The pH will vary. exchange alkali fromOion nit reg een removed flow passes through the plant of The Dow of the Chemic entire river before continuing down the river. In any case ompany corrosive and willkend to carry sludge.

the pond several times (b) Water supply for ion exchange.

water will be It is intended initially to use Midland city water (Hur serve as well) as makeup, and after the first ~ year to on use wateri would water 'from The Dow Chemical Company. The Saginaw-Midland on exchange W takesand water from a crib140 several a er H Authority Point pumps it about miles.miles outinterce The crib in uronLake att Whitestone current that is low in impurities because it comes fromp s a Lake Superior The Dow Chemical Company uses ~ raw Huron watera ingranite se bed.

According.to the Midland water plant laboratory veral operations.

water have much the same typical composition , city water and Huron Ions Huron , differing chiefly in pH City Na ,+, 2 ppm.

Ca 4 ppm.

Mg*+ 26 8 30 Cl- 8 12 s0g-- 15 30 20 pH 79 9 0-9.4 The Huron and purification, water fluois _ridated.

limed to reduce hardness

, treated with chlorine for (c) Cation exchange water.

The Dow Chemical operating on Company passes Huron water them with Na the Na+ cycle. This removes alhrough a cation exchanger heavy metal ions, replacing ions. The exhausted exchanger is regenerated with Nacl solution Huron and water onewashed.

might expect:From a cation exchanger on the Na+ cycl e fed with Na*

Cl- 28 ppm.

s0 -- 14 3 25 pH (d) Hydrogen ion cycle. 8.6-9.1 first step cation exchanger must be operated e water, theon the H it must be regenerated with strong acid, for example with H cycle, that is, 250g, rather

e than with salt, and wash:d. One might expect from Huron water:

hcl 14~.5 ppm.

H250 g 25.6 pH 6, approximately When.a cation exchanger of either type is exhausted, Na ,, Ca ,,, etc.

ions will pass through, and it must be regenerated. It is the commonest thing in the world for operators to permit this to occur.

The output of acids from the operating exchanger depends upon the anions present: HC1, H 2S03 and H 2003 for example.Different resins are used for weak and for strong acids (e) Anion exchange water.

An anion exchanger does .not exchange ions, but, instead, absorbs acids

~

(or if intended for bases, it absorbs bases) as such. The product is essentially pure water. When the exchanger is exhausted, acids pass through. This results in a drop in pH and an immediate increase in specific conductance. However,it is common for operators to not notice this. Some specific conductances of ions at 25 C. include:

H* 349.8 C1- 76 3 Na+ 50.1 OH- 198.0' iCa++ 59 5 iS0 --

g 79 8 The specific conductance due to hcl is thus, for example, about 3 4 times that due to Nacl.

The anion exchanger is regenerated with NaOH or with Na 00 .

(f) Other considerations. 2 3 If . city water is used, it will contain chlorine, which is not removed by an exchanger. The fate of F- ions is uncertain.

Water, consisting of returned condensate and fresh water makeup, will be demineralized and stored. It will be introduced into the secondary cycle, which also contains so-called " full-flow polishing" demineralizers. These, too, can exhaust and pass cations and acids. Ions passed will accumulate in the steam generator and can only be removed by blowdown. Meanwhile they will corrode the equipment.

Ion exchange resins can be subject to cracking of the beads and to et;osion. Eventually resin will enter the steam generator, whether due to ens ,osion or to excessive velocity of'the feed water. No mention is made of a polishing filter to retain resin. Cation resi.ns contain acidic groups. Anion resins contain basic groups. If cation resin enters the anio&(Achanger, the result is uncertain. Resins must be added

! or replaced after long usage.

(g)0ther chemistry.

l It is stated that ammonia will be used for pH control and hydrazine for

! scavenging oxygen. NUREG 0571 says that hydrogen will be used.

i Ammonium chloride is very corrosive.It hydrolyves to an acid pH. Hydrazine hydrochloride reduce corrosion.

or sulfate will be corrosive. Silicate, if present, might

, 7. The steam generator. -

l

' The Babcock and Wilcox once-through steam generator has primary coolant on the tube side and secondary coolant on the shell side. The secondary coolant is preheated in the annulus and converted to steam in the shell.

NUREG 0571 states that the tub'es are of Inconel, 0.625 inch 0.D., 0.035 l

inch wallthickness. If one a umes a factor of safety of 6, the working pressure of the tubes is abou ,500 psig.

i

. 5 Inconel, a product of International Nickel Company ,

contains:

Nickel Chromium 79 5T=

Iron 13.0 C,arbon 65 Copper 0.08 Manganese 0.20 unknown 0.25 Of these components, copper is 0.47 certain whether Inconel is a developed alloy, or whetherUrpesireable; it is a316 steel composition found in natural ore. Inconel resembles , like Monel,stainl in many propqrties. ess Inconel HC1, to Hhas PO only partial resistance to NH 3 Cl, to Ca(OC1')2, to C1 , to 3 2 according to Perry's Chemical Engineer's Handbook.and 3 to Ca(

Tube failures system. have occured when phosphates were used a y initi in the ll in order to prevent vaporization in the reactor o be 2 000(a)PSIG.,Press produced at 1,000 psig., or about 555 F.the genBrator tubes relationship for water includes: The pres,sure temperatureg. Steam is t Temp. OF 555 Pressure, psia.

600 1015 640 1543 700 2345 '

3093 705 4(c) 3206.2(c) is questionable rise whether to 705.4 F., the critical point, pump pressure . could to 6400F.,

e . Should beit ma Tube failure in the steam generator could occur forachemic or for lthe rea mechanical wo.uld be 1,000reasons.

psig. The driving forse for leakage to the ' seconda e (b) The corrosion resistance of metals such as nickel and depends upon theisformation of a passive film of oxidestainless or the liksteel the surface. This initially es e upon alkaline conditions. In an oxygentablished by oxygen and promoted by free atmosphere if it is lost, it cannot (c) Whenbe reestablished. Chloride and sulfate ions, interfere with dissimilar sv pasy.

anic couple is formed and a potential set up. The metal at potential can occurwi 11 be the between anodemetals, dissimilar and will tend to go into solution . Couples metals that have experienced dissimilar stress. ween copper, Inc for example. Copper not only sets up a couple, but it is di sselved as a complex by ammonia. A welding rod of the wrong composition , such as granular structure, corrosion can begin at the crystals. ra interf A few single electrode potentials include:

Fe/FeSO -0.44 h

Ni/NiSO -0.23 Cu/CuSO4 t0 34

6 -

(d) The water in the primary cyclo will contain impurities that occur in the initial feed water, or in water introduced to replace leakage and - _

blowdown. Besides the initial components of the feed water, it can contain metallic components of th e reactor, the generator, the piping and the pump. I have not seen a stuffing box nor mechanical seal on a pump that will not leak sooner or later. Gases and vapor will collect above the tube sheet in the generator.

(e) Secondary cycle.

As one unit i,s to produce steam for sale, the secondary cycle of that u-nit will be more subject to corrosion than will that of the other unit. There will be a. buildup of ions from the feed water and from ion exchange operations, and from chemical treatment acids, chlorine, cations, anions, air and contaminants from the materials of construct-ion.These will be removed by blowdown. The shell side of the steam generator is stated to contain iron baffles as spacers and supports.

These and the shell are subject to corrosion and to electrolysis.

There is no mention of attached anodtigSludge will collect on baffles, supports and the lower tube sheet.As steam is removed, these will concentrate, and blowdown will be necessary. It is essential that a11 construction debris shall have been removed. The driving force for leakage will be 1,000 psig. A critical point of corrosion will be at the liquid-vapor interface, and in the area above this where entrainment occurs.The vapor velocity will be much greater than the liquid velocity.

6- The condenser.

Hothing is said of the ccnstruction of the condenser, but for operation at perhaps 20 inches of rercury vacuum, 190 F and with nominally salt-free water on the tube side, it may be assumed to be of conventional iron construction. Corrosion products will develop in the shell, and remain there, or pass into the hot-well.

The tube sheets and tubes will be corroded and sludged up by components in the co611ng pond. In~ case of a leak, these will be shaked into the shell and pass into the hot-well, from where they will enter the second-ary water system.The driving force for leakage will be about $0 psig.

Fouling of this condenser will be extensive, but, aside from leaks, the chief danger to the secondary sater supply lies in the uncertainties of ion exchange operation and chemical control, and improper materials of construction. ,

Admittedly, 3400 tons per year of sulfuric acid, presumably 66 Be.,

or 775 pounds per hour will.be used to regenerate the cation exchangers and will enter the pond, partly neutralized.

9 , Mechanical.

The Babcock and Wilcox steam generator differs from others in having two rigid tube sheets, rather than hairpin tubes fixed to a single sheet.It should be relatively free from vibrational stress and accomp-anying wear, but suffers from lack of provision for thermal expansion.

A floating tube sheet, occasionally used, is not practical. If the tubes are straight and rigid at room temperature, they will expand and be bowed at operating temperature. Conversely, i'f they are rigid and straight at operating temperature, they will be drawn at normal temperature, perhaps beyond their proportional limit. It is not stated, but the tube sheets are probably drilled. Boring, or broaching (in the sense of burnish-ing) would give smoother contact surfaces. When the tubes are expanded to fit the holes in the tube sheet, the stress exceeds the proportional limit and thereafter they are liable to further expansion under stress.

If the tubes are e,xpanded only on the outer side, crevices will surround the tubes on the shell side and corrosive" components can build up .

If the tubes are expanded through to the shell side, it makes their removal more difficult and strains extend further into the shell. In either case the stress on a slender column under axial load is added

._ 7 to the stress of Pressure. Stresses will initiate intergranular corrosion of the tubing.

Sheuld a tube fail, it would be a cantilever with one fixed end. It would vibrate and rub, both against other tubes and against the tube sup orts,although not as much as in other designs.

The Babcock and Wilcox tube spacers are broached, rather than drilled.

If by broaching is meant burnishing, they would present fairly smooth contact surfaces.

NUREG 0571 indicates tube failure due to fatigue near the top of the generator, that is, near the liquid-vapor interface or in the vapor space.

plus Stress column is caused action by vibration and corrosion of the tubes in the vapor space at the interface.

Foreign material can collect at the tube sheets and supports, and could wear the tubes. Construction residues must be completely removed.

Ihe tubes can be dented by foreign bodies, or thinned by stress and corrosion.

errode nearbyIf faulty tubes.tubes are plugged, the resulting flow currents may So-called " sleeving", that is the placement of a piece of smaller tube inside of a stressed, tube at the tube sheet, and expansion to provide tightness, is not a desirable procedure. It further stresses the tube and increases velocity through the sleeved area, and may cause cross ,

currents above the tube sheet.

10. Other comments.

In the case of generator tube rupture, primary cooling water will leak into the secondary system at 1,000 psig. driving pressure, resulting in some repair willloss be of cooling and introducing some radioactivity. Shutdown for necessary.

It is obvious that the use of pond water for eme gency cooling will contaminate both the reactor system and the pond.

Clearly the best way to cool a reactor would be to lower the control rods, to-keep steam. the liquid Steam removal levelprovide would above2the rods and to remove vapori7ed of water times the heat removal per pound that liquid cooling wood, as a minimum. NUREG 0916 indicates that the aplomb of the operators in the Ginna plant tube failure event resembled that of a cat on a hot stove, with.little attention to engineering principles, and continuous indecision.

Steam evolved in an emerfency could be directly condensed with water sprays in a suitable vessel and the inert gases vented.

A serious source of atmospheric contamination is the use of pop valves for reseat relief. Such valves are subject to wire-drawing and generally do not properly.

ihey are unsatisfactory for use where toxic vapors and gases are involved. Parallel frangible safeties, with their stems l

locked much together so that if one accessible, the other is not, would be safer.

The air-operated venting valve in the primary. system is inappropriate.

First, it can be valved off by a blocking valve, which would make it ineffective and might cause transients if closed suddenly, and, second, if it is direct acting, air failure would close it, even if needed, and, if reverse acting, air failure would ppen it, equally bad. Closure of valves to the turbine would cause transients. causing unwante Finally, the discussions of steam. tube integrity and the studies at Brookhaven and the Franklin Institute are excellent.

.... 8 -

My name is A.B. Savage. I am a retired.Dow Chemical Company research engineer. I do not speak for the company,,nor have I used its facilities in this matter.

I attended.the University of Minnesota and received the degree of Bachelor of Chemical Engineering with Distinction in 1935. Courses of special relevance included resistance of materials, metallography, industrial electrochemistry, electric power and mechanical engineering.

I was elected to Tau Beta Pi, honorary engineering society,and to Phi- Lambda Upsilon, honorary chemistry society.

I had a fellowship with and worked in the plant of the Minnesota and Ontario Paper Company, now owned by Boise, Cascade, 1935-1937. I received the Degree of M.S. in Ch.E. and was elected to Sigma Xi, honorary research society. Experience included reactions involving corrosive acids under pressure, pH control and the observation of evaporator operation and maintenance.

I worked as a research chemical engineer-under various titles for the Dow Chemical Company for 39t years. I was an active me.mber of the American Institute of Chemical Engineers and the American Chemical Socity. I was elected to the Research Society of America. Experience included operation of reactors at about 500 F, operation of reactors at about 200 psig.,cperation of ion exchangers, water quality, pH, conductometric and potentiometric measurements, removal of salts from products and regulating their pH, testing for corrosion, and design and operation of continuous and batch stills, condensers and heat exchahgers, and operation of a process involving corrosive acid under pressure. I also had to do with prevention of possible runaway reactions, and many aspects of safety, including frangible safeties, relief valves, etc.

I took a course in continuous reactor design from a University of Michigan instructor.

I hold m. ore than 40 U.S. or foreign patents and have, written technical book chapters and encyclopaedia articles.

.