Oxygen Analysis Control in the Open Hearthlibrary.aimehq.org/library/books/Open Hearth 1958/Open Hearth 1958... · The use of the standard Orsat in waste-gas analysis work has been

OPERATING AND COXBUSTION 461

not made any changes. We are still in the process of duplicating operating furnace results in the model furnace. This is, of necessity, the first step. I think the next step in the program will be to go ahead with some changes of design and evaluate the effects on flow distributions. Such changes can be made and the results analyzed much more easily and economically on the model than on the operating furnace.

F. J. TODD-&&. Brion, in your operating furnace, was the air forced or induced?

D. F. BRION-Forced air in combination with induced draft.

F. J. Tom-Induced draft afterward. In downtakes, might the difference between the model and the operating furnace have some relation to that?

D. F. BRION-I do not feel that it makes any dserence whether the air is blown or

sucked through the model. (Experimental results obtained since the meeting have proved the point that the distribution of combustion air in the uptakes and waste gas in the downtakes remains unchanged when induced draft is substituted for forced air.)

J. H. RICHARDS-This question has arisen concerning water-model studies also. We have both pushed and pulled water through our slice modeIs of checkers and have observed no change in the flow patterns.

E. C. MCDONALD, CHARMAN-Are there any other comments? If not, we will go on to the paper by Mr. E. W. Hunziker, Division Superintendent Open Hearth and Foundry, and Mr. J. W. Bain, Fuel Engi- neer, Open Hearth and Foundry, Colum- bia-Geneva Steel Division, United States Steel Corporation, Geneva, Utah. Mr. Hunziker will present the paper.

Continuous Oxygen Analysis for Combustion Control in the Open Hearth BY E. W. HUNZIKER AND J. W. BAIN

FURNACE operators have long realized the importance of supplying the proper amount of air to bum the fuel fired in the open hearth furnace. They know that to operate an open hearth furnace with a de- ficiency of combustion air tends to shorten refractory life and cause an increase in carry-over of unburned fuel into the checkers, which will gradually slow the furnace operation as the campaign prog- resses. When an excess of combustion air is delivered to the furnace, it increases thc fuel rate by decreasing the flame tempera- ture and increasing heat loss to the stack. For good open hearth furnace operation, the control of excess air is desirable in the interest of fuel economy and maximum furnace productivity.

Many of the modern open hearth fur- naces are equipped with fuel-air ratio

control, which is accomplished by means of volumetric flow measurements on both the fuel and combustion air. The fuel-air ratio control is subject to many variables; for instance, changes in specific gravity, temperature, pressure, Btu value of the fuel being used, and changing chemical reac- tions. I n addition, the open hearth furnace presents one other important variable, in that it must be assumed by the ratio con- trol that the same volume of air is being delivered to each end of the furnace. This is not likely, as the air infiltration in flues, checkers, fantails, slag pockets, and up- takes often varies from end to end over a wide range. If this infiltrated air can be held down to a small percentage of the combustion air being supplied, reasonable control of fuel-air can be maintained.

Considerable time and effort have been

46 2 PROCEEDINGS OF OPEN HEARTH CONFERENCE, 1958

used in determining the products of corn- No. 8 furnace early in May. Both analyzers bustion of waste gases in the open hearth have operated with negligible outage time furnace to establish the correct firing prac- and minimum maintenance. Compared tice with respect to fuel and air present. with the other furnaces in the shop, the

The use of the standard Orsat in waste-gas analysis work has been of great value; however, this method of spot analysis taken during the meltdown does not take into account the changing conditions that arise during the remaining part of the heat. The results obtained from waste-gas analysis show that COz and 0 2 are both indicative of combustion efficiency in the open hearth furnace. As a means of combustion control, only the 0 2 is reliable, as the CO, content in the waste gases varies with the type of fuel being used and from CO emission from the bath. Maximum combustion efficiency is indicated by maximum percentage of Cop, but maximum percentage of C 0 2 will decrease with either a decrease or an in- crease of combustion air.

Early in January 1957, Geneva Works open hearth department installed a con- tinuous waste-gas analyzer on No. 6 open hearth furnace. The first month's results so far exceeded expectations that a second unit was ordered and put in service on

production rates are higher and fuel con- sumption is lower. The original plan was to completely eliminate the existing fuel- air ratio control and to directly set com- bustion air flow from waste-gas analysis. This so-called "end point control" has operated most satisfactorily with oxygen content in the waste gases controlled to very close tolerances.

The oxygen analysis equipment consists of the sampling system, sample conditioner, analyzer, and oxygen recorder furnished by Leeds and Northrup. A water-cooled stain- less-steel probe (Fig I) is inserted in each downtake of the furnace about 24 in. above floor level in the port end. I n the corner buckstay of the furnace are mounted the other components of the sampling system, consisting of water and steam connections, steam ejector, pressure gauges and regu- lators, condenser, and drain line (Fig 2).

T o maintain probe opening, a spray clean- ing nozzle inside the probe directs water through three small openings, a t a 45-deg angle in the opposite direction of sample flow, about one inch from the probe tip. The steam ejector aspirates the waste-gas sample through the probe to a point just upstream from the condenser where cooling water is added to the steam-sample mix- ture. The water is dropped out to drain a t the condenser through a double-seated float valve and the clean sample leaves the top of the condenser under 6 to 8 lb pressure, from which it travels through >din . copper tubing lines to the analyzer in the control house, across the charging floor. I n the con- trol house, near the analyzer, are two sole- noid valves (Fig 3) connected to the furnace reversal system, which direct the sample from the proper end of the furnace to the analyzer (Fig 4). An air cooling system around the common sample line serves to keep the sample temperature below dew

OPERATING AND COMBCSTION 463

point before i t enters the sample condi- analyzer can be made on air or bottled gas tioner and regulating unit. .4t this point, of known oxygen content, and a zero check 0.5 dh of the go-dh sample drawn from can be made by disengaging the magnet the waste-gas stream is transferred by from the measuring and reference cells.

means of small-diameter plastic tubing through a filter and rotameter to the cells of the analyzer. The analyzer cabinet is thermostatically controlled a t q o ° F and contains the analyzing assembly function- ing by means of the paramagnetic property of oxygen. About 15 seconds is required for a change in waste-gas oxygen content a t the furnace to begin showing on the oxygen recorder. Calibration checks on the

Changes in calibration are not often re- quired but can easily be made with two potcntiomcters mounted in the recorder case, one for the zero check and one for range. A double bridge circuit, one measur- ing and one reference, is utilized in the analyzing unit to offset any errors from pressure changes or natural thermal con- vection. The output from this circuit is transmitted to an electronic recorder

mounted above the analyzer, which has a time the controller calls for a change for range of o to ro pct oxygen. the recorder to pick up the change.

The oxygen recorder is equipped with The motor-operated jet-pipe adjuster the necessary components required to make feeds oil to the open or closed side of the

FIG ~--S.~JIPLI? RE\'ERSAL VALVES.

i t a controller and has a manual control setter, so that the oxygen content of the waste gas can be controlled within a very narrow range. Both hydraulic and electric systems have been used to establish com- bustion air flow to the furnace from the oxygen content of the waste gases. On the hydraulic system, after the waste gas is analyzed, the signal calling for an increase or a decrease in air flow is transmitted to a Leeds and Northrup series 60 position- adjusting type relay, from which the signal is sent to an Askania motor-operated jet- pipe adjuster (Fig s). The position-adjust- ing relay dampens and delays the signal to the jet-pipe adjuster by means of propor- tional band, rate time, reset rate, and repeat adjustments. This is important in order to prevent hunting and over control, as about forty seconds is required from the

hydraulic cylinder controlling the forced- draft fan louver position (Fig 6). This is a so-called cascade system, with an impulse feed back from the combustion air orifice to the diaphragm balancing one side of the jet pipe. The position-adjusting relay slide wire positions the slide wire on the motor operating the jet-pipe adjuster.

On the electric system, a Leeds and Northrup 1x0-volt drive motor attached directly to the forced-draft fan louvers operates from the output of the Leeds and Northrup position adjusting type relay (Fig 7). The settings on the relay adjust- ments necessarily are different for the all- electric system in order to obtain results comparable to the hydraulic system. The electric system employs a simple balance between the drive-motor slide wire and the relay slide wire.

OPERATING AND COMBUSTION 465

The oxygen analyzing and control system is completely separate from the fuel meas- uring system and has eliminated the fuel- air ratio control previously used (Fig 8). This method, rather than using the ana- lyzer as a trimmer on a fuel-air ratio con- trol, has accomplished two purposes: (I) the system is simpler and has fewer com- ponents to be serviced and maintained, and (2) better combustion control is maintained by using the anaIyzer as the al~soluto end point control.

Sample tube location has bccn changed from the original end wall to the port end of the clowntake (Fig I). for four reasons: ( I ) Orsat traverses of the downtake showed equally represrntative samples from either position, (2) the probes collected less dirt when inserted in the port than in the side end wall, (3) if the jet wash nozzle hurnetl out, the water would fall harmlcssly into the slag pocket instead of shooting across the downtakc, striking the suspendcd brickwork on the other side, and (4) the location a t the back of the furnace exposed the sample probe to less congestion, traftic, and possibilities of damage from equipment or workmen.

F ~ N A C E PRESSURE AND INFILTRATION

Furnace pressure control is of vital im- portance in maintaining good combustion from continuous waste-gas analysis. Pres- sure must be uniform on both furnace ends and should be about 0.08 in. H a . Too little furnace pressure permits excessive upstairs infiltration of cold air, which, owing to fast response and close control, cuts back on the preheated air furnished, with a result- an t loss in furnace efficiency. Excessive furnace pressure, of course, is detrimental to refractory life.

Infiltration problems have always been a source of poor furnace operation. A differ- ential in each furnace end can be handled

FIG 4--PANEL SHOWING KECOBDER-C ROLLER, OXYGEN ANAT-YZER, POSITION IUSTLVG TYPE RELAY.

ION- AD-

easily and adequately by continuous waste- Geneva Steel Division (Fig 11). For both gas analysis where proper compensation tests, the oxygen analyzer was taken off cannot be made with a conventional fuel- control and the recorder was covered and air ratio control. Differential in measured locked. By this means, a 24-hr record of

combustion air flow fromeach furnace end, as controlled by the analyzer, also quickly points to a loose furnace end or an end where brickwork may have fallen in. By continuous oxygen analysis of the waste gas, the furnace runs more evenly and both ends operate a t the same rate, even though a differential in infiltration exists between ends. Fig 9 sho~vs the difference in combustion air furnished to each end while uniform oxygen content is main- tained in the waste gases.

waste-gas oxygen content was obtained while the furnace operator had only the normal combustion tools a t his disposal. The furnace operator was asked to use his own judgment for the manual combus- tion air settings test. For the fuel-air ratio test, air was set from the amount of fuel fired, as is normal for fuel-air ratio control.

The statistics in Table I show the first 9 months of operation on No. 6 furnace a s compared with shop averages. Table 2

shows temperature da ta taken on the preheated air and regenerative system of

ADVANTAGES OF CONTINUOUS ANALYSIS furnace No. 6, In order to show more graphically th.e I n October 1957, Geneva Works received

advantages of continuous waste-gas analy- approval for completely equipping its sis, two tests were run to find the results xo-furnace shop with waste-gas analyzers that can be expected from manual com- on the basis of the results that had been bustion air control (Fig 10) and from the obtained on No. 6 and No. 8 furnaces. It fuel-air ratio control used a t Columbia- is expected that this program will be com-

OPERATING A N D COivIBUSTION 467

plete by June 1958, and will show the E. C. WICDONALD, CHAIRMAN-Are improvements on all the furnaces that there any questions? have been shown to date on No. 6 and JOHN DICKSON-We have a similar in- No. 8. The benefits that will be realized strument in our Wheeling Steel shop al

from an installation of this type are tangi- Steubenville, Ohio for recording only. You ble, and if the equipment is properly used have done a fine piece of work, Mr. and maintained, reduced fuel rates, in- Hunziker. Will you dcscribe your main- creased production rates, and better tenance time? Do you use steam tracers, refractory life should be obtained. or what precautions are used on the

sampling lines for protection against cold

DISCUSSION weather?

The pattcrns on your chart showing the E. W. HUNZIKER-If there are any ques- fuel ratio control are very similar to the

tions here, I want to introduce the fellow patterns we have obtained in our fuel-air who does the work, Mr. J. N. Bain. We ratio control shop. Assuming that your cannot let him off without doing somethinq hot-metal practices were 55 to 60 pct, do anyway, can we? you ever experience bath freezing because

468 PROCEEDINGS OF OPEN HEARTH CONFERENCE, 1955

FIG 7-ELECTRIC-DRIVE XOTOR ON COJiBUSTION AIR F A X .

TABLE I--Operaling Dale, No. 6 Furnace vs Shop Average

Period Avg W t Avg T a p Avg Chg. Tons pcr T o n s per MhIBtu per 1 I per Heat I t o Tap 1 to Tap 1 01. Hr / M. Hr 1 Net Ton

January 1 9 5 7 . .

February 1 9 5 7 . . .

March 1 9 5 7 . . . . . . . . . . . .

April 1 9 5 7 . . . . . . . . . . .

May 1 9 5 7 . . . . . . . . .

January through May 1 9 5 7 .

September 1 9 5 7 . . . . . . . . . . .

No. 6 Shop

No. 6 Shop

No. 6 Shop

No. 6 Shop

No. 6 Shop

No. 6 Avg Shop Avg No. 6 Shop

Period

January through May, 1 9 5 7 . . . . . . .

Shop Average

1 1 8 . 2 7

102

Roof and Front-Wall Life

Roof life Average number of heats Front wall life Average number of beats

No. 6 Furnace

1 3 6

119

OPERATING AND CO&CBUSTION

FIG 8-OXYGEN Water-cooled sample probe. Sampling station (steam ejector). Solenoid reversing valve for sample. Oxygen recorder-controller. 34agnetic oxygen analyzer. Position adjusting type rciay. ktotor-operated adjuster.

ANALYZING AND CONTROL SYSTEM. 8. Jet pipe. 9. Ihaphragm. 10. Crank-type cylinder. I I . Porced-draft fan louvers. 12. Combustion air orifice. 7, 8, 9,

tlraulic componenls. 13. Klectric-drive motor.

TO. Hy-

Temperature North Checker, Deg F Test.

So.

I

Temperature South Checker, Deg F

Uptake

2940

2860 2870

2955

From Start of Furnace.

D~~~

38 39 4 6 5 8 65 7 4

128 292

Temperature of Air, Deg F

Uptake

2880

2875 2940

2760

Bridgewall

2420

2320 2390

2410

Beginning of Cycle

2705 2695 2665 2670 a670 2675 2860 2360

Bridgewall

2435

2320 2430

ZZgO

End of Cycle

252.5 2560 2525 2580 2560 2570 2795 2175

OXYGEN CONTEh

/ V i O h

A'-

I T OF WAS

ic>tr- - ' 1- TE GAS

.

t , , woo*

OMBUSTION AIR FLOV

FIG 9-AIR INFILTRATION EFFECTS ON COHB~BGSTION AIR SDPPLY.

OPERATING AND COMBUSTION 4f1

-- - . ,- - -- . *- .FT 5- . - -

-4 k

\

OXYGLN CONTENT OF WASTE GAS COMBUSTWN AIR FLOW

FUEL FLOW

FIG 10-OXYGEN COXTENT IN WASTE GASES WITH AIR AND FUEL CONTROLLED MANUALLY.

47* PROCEEDINGS OF O P E N ITEARTH CONFERENCE, 1958

OXYGEN CONTENT OF WASTE GAS COMBUSTION YR FLOW -- .

FUEL rrvn

FIG I 1-OXYGEN CONTENT IN WASTE GASES OBTAINED FROM FUEL-AIR R:VrIO CONTROL.

I OPERATING AND COWUSTION 413

COMBUSTON YR FLOW OXYGEN CONTENT OF A'ASTE GAS

eu- s- t

/ .. "\ / p ,

I. a \\

FUEL R O W

12-OXYGEN CONTENT IN I~IASTE GASES OBTAIKEU F K O ~ CONTROL RY CONTI>TOUS OXYGYK ANALYSIS.

of the necessarily high amounts of air similar in nature* where the analyzer is required to neutralize the furnace atmos- used for the control of the fuel rather than phere during the more violent fume pe- combustion air, I should like to know some riods? The control of 5ue-gas oxygen, I of the conditions that must be handled in believe, is good practice where there are your shop, as well as your reasons for little air or flue limitations in the furnace deciding on the control of combustion air system and plugging conditions are not rather than fuel. . - - -

critical. However, with the overexpanded furnace laboratories on the smaller or original downstairs, the flue capacities are somewhat strained and would prohibit the use of end-point control. Would you care to comment on this last statement?

J. W. BAIN-Our equipment is checked every turn. The instrument is the responsi- bility of the instrument department; we have not hired any more men to take care of this equipment. I t is checked every turn for pressures, sample pressures, and water pressure. Maintenance time, I would imagine, is about I hr or 134 hr per week.

Our sample lines are steam traced within about 2 0 f t of the control panel.

We have not experienced any bath freez- ing on these furnaces.

I cannot answer your fourth question- I do not know.

JOHN DICKSON-What I had in mind was that, once, a good while ago, when we did not have this installation, we were using portables, and checking through entire heats, and we happened to catch one heat that had a very violent fume period due to charging time and so forth. We - -

found that as the limit of fuel-air ratio was approached (52 pct excess air), we still could not kill all the CO. Also, freezing started.

There again we may get into overex- panded upstairs and small downstairs, in that we were limited as to fuel. Tha t is the reason I hoped to get a comment on this.

J. W. BAIN-I do not know.

K. A. ELXAN-With reference to the paper this morning which was somewhat

J. W. BAIN-W~ use a firing profile. We like to maintain a Btu a t the open hearth setup. We start out with 138 million Btu during the meltdown, and during the latter part of the flush off, we cut the fuel back to 112 million. During the lime boil, we follow the practice of go million, and during the working period heat, we step it back up to 112 million. We have sufficient com- bustion air to handle all of the actions that take place. Does that answer your question?

K. A. ELKAN-Mr. Bain, you say that you do have an unlimited quantity of com- bustion air to handle the problems. Why did you decide to vary the larger volume fuel if air can be considered as a fuel rather than the fuel itself?

J. W. BAIN-Our maximum air flow is 2 million cu ft, and we do not have very much trouble handling it.

K. A. ELKAN-What speed of response do you get by changing combustion air, which has to travel through the checkers, fantails, and uptakes, as compared with changing the fuel, which is controlled a t the end of the furnace?

J. W. B ~ ~ ~ - N o r m a l l y , from the time you change the air a t the controller, moving the combustion fan louver up to the other side of the furnace, it is about 40 seconds.

L. F. REINARTZ-HOW often do you find it necessary to clean the sample tubes, and how much difficulty do you have in cleaning the tubes?

* McElhaney. C. J., and W. V. McNiece: Continuous Waste-Gas Analysis for Automatic Fuel Control in the Open Hearth. This volume, page 352.

O P E R A T I N G AN 'D COMBUSTION 475

What percentage of oxygen in the waste gases do you set as a standard? Do you vary that percentage throughout the cycle of the heat?

J. W. B A I N - - - ~ U ~ sample tubes are cleaned once a week. We run a constant 2 pct oxygen in the waste gases.

B. M. LARSEN-This system is prcsuma- bly flexible enough so that if you wished, you could vary the excess oxygen to favor bath oxidation.

B. M. LARSES-And also, if you wanted, you could operate both fuel and air by combination of the roof control and the oxygcn control system. Has that been tried out?

J. W. BAIN-That is our next project

M. S A ~ L E - W e are somc\vhaL sympa- thetic with this idea. We have purchased one of these analyzers and have installed it in one furnace, but we have run into a few difficulties.

We started with the zero to 5 pct scale and I noticed that your instrument had zero to 10. Our zero to 5 pct scale gave us a spread of about half the chart. The furnace normally breathes in and out and the oxygen analyzer would run from maybe 2.5 up to 3 or 3.5 pct, and swing back down to 2. On your control, how much swing, how much variation is there in the oxygen content during the whole reversal? I t cer-

tainly does not hold one constant per- centage, I should think.

J. W. BAIN-So, the first two analyzers recordings we purchasrd were zero to ro. The eight additional ones are zero to 5. Thc control band is about half as wide as that we show on this other recorder, if that is what you mcan.

E. C. MCDONALD, CHAIRMAN-Are there any othcr commcnts? If not, that concludes the meeting. I think Mr. Yarotsky has somcthing to say.

M. F. YAROTSKY-AS thc curtain falls on the technical portion of the 1958 con- vention, we can honestly appraise i t as one that has bccn kcpt to the high standards we expect from our technical sessions. 3luch of thc credit for that goes to C. &I. Kay and his committee for developing the program and rnlisting support of partici- pants. We cannot help but leave the con- vention wiscr through learning of what has been accomplished by our collcagues in business. ,\ splendid job has been done by all participants, for which we are grateful.

Arrangements are underway for our next national mccting, early in April in 1959 I am extremely proud of the work our or- ganization is doing, but in order to con- tinue it successfully, i t needs our sustained support both mentally, spiritually, and financially. Gentlemen, the meeting stands adjourned.

[The meeting adjourned a1 4:30 P.m.]