PERIODICA POLYTECHNICA SER. ME CH. ENG. VOL. 37, NO. 2, PP. 103-110 (1993) EFFECT OF THE OIL ADDITIVES ON THE THERMAL POvVER STATIONS AND THEIR AIR POLLUTANT EMISSION Attila MEGGYES Department f Heat Engines Technical University f Budapest H-1521 Budapest, Hungary Tel.: 166 4011/2621 Fax: 1664679 E-mail: meggyes@kalrgep.bme.hu Abstract On the basis f the facts summarized abve, the applicatin f the additives is ratinal, when the technlgy f supplement can be supervised. And the recvery f dust cntent frm the flue gas has been elucidated, therwise the disadva.nta.geus ffset the benefits expected. Keywrds: pwer statins, il additives; NO x emissin, 802 emissin. Intrductin The bilers in the hydrcarbn fuelled pwer statins perate in wide range f their capacity, which increases the crrsin n the flue gas side, extremely at the case f fuel ils. The crrsin is increased by disadvantageus change in the cmpsitin f the fuel (higher sulfur, vanadium, and asphalt cntent, lwer calrific value) [1]. T decrease the undesirable effects mentined abve, we have tested different il additives in Danube Pwer Statin C. fr several years. 1. Advantage and Cmpsitin f the Additives The additives must cmply with the fllwing advantages (as prducers prmte these prducts): - increase ph value f the fly ash abve 4, decrease in the FeS04 cntent f the fly ash significantly (at least by 50 %) (declining f the lw-temperature crrsin), decrease in the high-temperature vanadium crrsin substantially (at least by 50%), reductin in the temperature f the flue gas significantly (at least by 20 C) at stable ph value" 4.5-5.5 (higher efficiency), mre perfect cmbustin, less st emissin.
104 A. MEGGYES First f all, the prducers want t achieve the effects abve mentined, using MgO type additives which are mdified by different metallic xides. The additives cntain magnesium-xide in suspensin frm (density '" 1.7g/cm 3, MgO cntent", 1200 g/l, carrier - light fuel il). Table 1 shws the cmpsitin and size distributin f the three different additives (A, B and C). Table 1 Prperties f the additives Cmpnents A B C MgO% 92.1 92.7 81.8 CaO% 0.71 1.67 3.64 Si02% 0.52 0.88 4.39 Fe203 0.78 0.28 1.95 Al203 0.27 0.27 0.59 Diameter Average IJ.lm 7J.lm 5J.lm belw 20 J.Lm 100% 92% belw 10 J.Lm 100% 66% 71% belw 5 J.Lm 98% 43% belw 1 J.Lm 55% 25% 28% This table shws that all additives cntain MgO, as agent, at least 80% (tgether with CaO 85%). The size f particle is decisively belw 10 j.lm, even belw 1 j.lm, that means very small particle diameters. The small size MgO particles are expected t take part cmpletely in reactin with 803. 2. Results 2.1. Analysis f Advantages Ezpected Our first test was carried ut with additive B during mre than three mnths (between February 24 and June 9). At the beginning, the additives were mixed directly int the fuel (till March 17), later they were separately led int the upper part f the fire zne. As effect f additive, we gained the fllwing experience: 1. ph values started frm 2 and achieved 3.5 while additives were directly mixed int the fuel il (till March 17). After feeding mdificatin, the ph values increased quickly, but after cntaminatin f verheating
EFFECT OF THE OIL ADDITIVES ON THE THERMAL POWER STATIONS 105 system (TH4) drpped under 4. In the secnd step, after washing and cleaning, ph values increased rapidly and became steady between 4 and 5. 2. By reductin f the irn-sulphate, the irn lss in the Ljungstrm preheater, and the lw-temperature crrsin can be cntrlled indirectly. On the basis f the tests [2], the fly ash irn cntent declined frm 16% t 1.5-4.5%, and stabilized, which means 85% reductin. The weight cmparisn f test cassettes lcated n the cld side f Ljungstrm preheater cnfirms significant reductin in irn lss. Accrding t this weight cmparisn, the irn lss decreased by 62.5% as effect f using additives. 3. Decrease f the high-temperature crrsin was analyzed indirectly [2]. Melting pint f depsits cllected (mechanica.lly) frm verheat er tubes (TH4) was bserved. Accrding t these measurements, sftening pint f the middle layer, which was the heaviest, increased by 200-300 c, insuring substantial decrease in crrsin. 4. Dew pint declined frm 150-155 c (at 215 MW) t abut 140 C in the first term f testing. Althugh it gradually increased with grwth f the depsitin in the heating surfaces t 145 C. In the secnd term, after cleaning, dew pint fell t 130 C, indicating the lack f catalysis initialized by the sediment frmed n the verheater. Acidic dew pints were measured bth befre and after the Ljungstrm preheater. Typically the latter was lwer by I-2 C. During the testing perid, the cmpsitin f the fuel il was mre r less stable. Sulfur cntent was between 2.76-3.18%, vanadium 130-150 mg/kg, sdium cntent varied frm 2.5 t 3.0 mg/kg. 5. Rati f unburned particles was 20-30 m/m% in the fly ash at the beginning. Its value did nt change cnsiderably after using additives. Even if we realized a sma.ll degree deteriratin taking int accunt the excess vlume f the fly ash. 2.2. Analysis f Air Pllutants l41 During the representative measurements, we analyzed the air pllutant cntent f the flue gases. The air pllutants analyzed are: sulfur dixide (S02), - carbn mnxide (CO), - nitrus xide (NO x ) and - particles (dust). T measure gas emissins, we used devices wrked n the principle f infra red, and chemiluminescent detecting. Mnitring pints lcated after the air preheater (Ljungstrm), and befre the ECO (ecnmizer).
106 A. MEGGYES The change resulted by the supplement is shv'i"il n Figs. 1 and 2, as functin f time. The NO x emissin (Fig. 1) depends n the pwer and beynd that the amunt f the il additive. The additives d nt effect n the CO emissin, that is why it cannt be seen n the figures. At the beginning, the NO x emissin increased significantly (Fig. 1) n the basis f using il additive. Mainly this can be explained by the high additive-fule rati (1:1000). The additive was depsited n the surfaces in the fire space. This depsit restricted the heat transfer, which led t higher :flue gas temperature and frmed mre thermal nitrgen xide. When the additive rati decreased (1:2000 and 1:2500) and the biler was cleaned, the NO x emissin declined, whereas it remained abve the initial value. Table 2 shws the change f relative S02 emissin (S02 effjs02 max) as the effect f the supplement. Table 2 Relative 80 2 change during measurement Date SO? eff 80 2 max 24/02 0.90 0.90 02/03 0.91 0.91 17/03 0.87 0.87 07/04 0.86 0.86 09/06 0.86 0.86 Befre starting the use f additives (February 24 and March 2) all the sulfur cntent f the fuel was nt able t frm S02, the difference was abut 9-10% fund as S03. After supplement this range became higher (13-14%), but the :flue gas cntained nly 2% S03, the remnant (11-12%) presumably was in the :fly ash. The slid material - was nt measured directly, hwever, sme calculatins can be made knwing the cmpsitin f the additives. 1000 ml additive - accrding t the prducers brchure - cntains 1200 g MgO. If the rati is 1: 2000 (1 liter additive t 2000 kg fuel il), 1 kg fuel il cntains 0.6 g MgO. Burning 1 kg fuel il gives apprximately 10 m 3 dry :flue gas (lambda=l). 1 m 3 frm such gas includes 60 mg MgO. If the magnesium xide is entirely transfrmed int MgS04 (mlecule weight rati - 40:120) the :flue gas hlds 180 mg/m 3 MgS04. If 02 cntent f the :flue gas is 3%, it becmes 154 mg/m 3. The additinal slid material (dust particles) increases the dust emissin substantially (by 100%).
NOx r9/m ppm 7 1 539 I 700 1 11 ;t/omw j 600 SOOr026S -"--- ---l-i - hvl V ' ==--==---+-i I - /'00 30G 200 700 th513 24.02 02.03 n n 0 6-- x I 17.03 1: 1000 60 MW I I 5-- x- 07.04 1:2500 Fig. 1. NOx emissin as functin f time and pwer x 3% 02 x after air pre-heater befre ecnmizer 09.06 1:2000 Time Additiv () '"i." '"i [;; t:: :... tl tl... :z: :ll l'l :ll t;l t-." '"... --I
f2 1 91m3 0.20 5,712 ---it--i-----r------------... 00 j I I t8 (, I X'60 I. 01'0 t X160 0110 I 502 max t._ / 6 ql5, '28< I -"<0.,< t x,.''-'6s: X 60 I n..._._ 060 _. x '60 1 l )(60 100 160 \ X 60-01'O- - -r- _._._. I (110 502 eff x-. O,lOt2,856 ---t-t--+---+----------j 3% O 2 ;,. C} C} t;; In I 60-60 MW 160-160 MW 0,05 1,428 270-210 tw 24.02 UZ.U) 17.03 07.04 09.06 n n 1: 1000 1:2500 1:2000 Fig. 2. S02 emissin as functin f time x after air pre-heater befre ecnmizer Time Additiv
EFFECT OF THE OIL ADDITIVES ON THE THERMAL POWER STATIONS 109 2.3.1. Rle f supplement site 2.3. Analysis f Other Effects Unfrtunately, there are undesirable side effects during the supplement. Direct supplement and big dses result fast mdificatin f the heat transfer n the heating surface in the biler. At the beginning, additinal and fast depsitin was bserved n the previusly cntaminated surface in TH4. Parallel, when the rati was 1:1000, a part f MgO additive was depsited n the surface f the fire space clring it white and reduced the heat radiatin. Tw effects tgether resulted the verheating f the large surface reheater getting t much high temperature cmbustin prducts, which culd nt be balanced by injectin. Naturally, the abve mentined effect culd be bserved in the further stages. The flue gas temperature befre the ECO went up frm 460 C t "" 570 C, causing dangerus impact n the steel cnstructin f the dilatatin, and the flue channels. Similarly, the flue gas temperature rse nearly 100 C after the ECO. Because the flue gas temperature increased dangerusly, principle f the supplement had t be changed. After 12 day wrk f the biler, we stpped the unit and had the injectrs installed. In the new system, we culd cntrl the rati depending n the utlet. On the basis f experience, we can annunce that the supplement int the fire space is mre beneficial than mixing int the fuel. In that case the ph imprvement is faster and there is n temperature shift. Hwever, the injectin int the fire space is technically mre cmplicated. 2.3.2. Depsits Befre testing the inlet part f the verheater, the black depsit thickness was 2-4 mm, and its remval was difficult. The supplement resulted higher thickness (4-8 mm), but it was much mre easily t be dislcated. Befre testing, the ECO was cvered by 2-3 mm thick, dusty, very fine, and easily remvable depsit f thickness which increased t 3-4 mm befre cleaning, but remained easily remvable. Generally, the physical prperties f all depsits changed in every surfaces. They became mre lse, prus and easily remvable (by hand). 3. Summary, Cnclusins We have gained the experience written abve, using additive type 'B'. But the results are similar applying additive type 'A' [5]. It is understandable,
110 A. MEGGYES because the additives cntain decisively MgO, as agent. The effect f the additive: 1. lw temperature crrsin decreased, 2. high temperature crrsin decreased, 3. acidic dew pint f flue gas decreased, 4. cmbustin f fuel did nt change, 5. gaseus pllutant emissin (CO, NO:!:, 802) did nt change, if the supplement was made in the suitable amunt, and lcatin, 6. slid particle emissin (dust) increased by the slid material cntent f the additive, 7. amunt f depsits increased, but it became mre easily remvable. On the basis f the facts summarized abve, the applicatin f the additives is ratinal, when the technlgy f supplement can be supervised. And the recvery f dust cntent frm the flue gas has been elucidated, therwise the disadvantageus ffset the benefits expected. References 1. O'BRIEN, J. C.: Practical Experience with Applicatin f a Magnesium Oxide Suspensin t a 2500 MW Oil-fired Biler. Jurnal f the Institute f Energy. September 1982. pp. 115-127. 2. Analysis f Dew Pint Decreasing & Crrsin Additive Effectiveness. Research Reprt, Technical University f Budapest, Department f Chemical Technlgy, 1987. 3. Oil-Additive Researches n the Biler N. 10 in Danube Pwer Statin. DPS 'Calric Technical' Team Reprt. Szazhalmbatta, 1987. 4. Air Pllutant Analysis in DPS. Part 1. Research Reprt, Technical University f Budapest, Department f Heat Engines. Budapest, 1987. 5. Oil-Additive Researches in DPS. Research Reprt, Technical University f Budapest, Department f Heat Engines, 1989.