Factors affecting gas oils response towards low temperature additives

Transcription

1 IndianJournalofChemicalTechnology Vol.2, November1995, pp Factors affecting gas oils response towards low temperature additives MMMohamed,H H AbouEl Naga& MF El Meneir Research Centre, Misr Petroleum Company, P.O. Box 228, Cairo, Egypt Received2 November1994; accepted16 May1995 Gas oil samples with different physico-chemical characteristics have shown variable response towards low temperature additives. Statistical interpretation of results indicated that magnitude in pour point and cold filter plugging point (CFPP) depression is greatly affected by the gas oil distillation range variables final boiling point (FBP) and difference between temperature to recover 90 vol % from temperature to recover 20 vol %. Per cent weight of n:-paraffinswith carbon number range C20-C32 is also found with great impact on that response magnitude. The statistical models derived have supported the high degree of correlation between these gas oil variables towards low temperature additives. In an attempt to increase the gas oils production by decreasing the per cent recovery at 350 C to be lower than 80% indicated that the incorporation of low temperature additives, e.g., pour point depressants and flow improvers were not always giving the expected good response. While some gas oils such as regular winter grades show high response towards these additives, whereas summer grades or even other winter grade~ are hardly affected. Many differences in gas oils respons~ towards low temperature additives have not been investigated clearly. The reported works are related either to the change in their crude oil sources or to the differences in their distillation rangel-9 The distillates could be characterized in terms of distillation as narrow boiling, broad boiling and broad boiling with high final boiling pointlo The gas oils with broad distillation range have high response towards additivesll. In the present paper, a statisitcal.study has been carried out on gas oils response to~ards low. temperature additives to explore for the correlation between gas oils distillation range as well as chemical structure variables and their response towards these additives. Experimental Procedure Nine gas oil samples with different physicochemical characteristics (Table 1) were selected for this study. The crude source for these samples was almost constant. Two effective pour point depressants and flow improvers (additives A and B respectively) were incorporate~ into these gas oil samples at concentration of 3000 ppm, which is the concentration recommended by their suppliers. Structures of these additives are based on polymethacrylates 12 (additive A) and polysaccharides (additive B). The difference in gas oil pour point and cold filter plugging point as additive and neat samples (pour point and cold filter plugging point depression, /:i.p) were considered as parameters for the a~ditives efficiency. Such difference in pour points and cold filter plugging points were found to be within the range 6 and 21 C for both additives (Table 2). Gas oil samples were separated into different chemical groups via liquid solid chromatography technique using GC under the following conditions Column Packet with 1.5 ml x 4 mi-d-glass 3 vol. % on Diatamite Co mesh Operated at Column oven temperature programmed looc/min C at Injector temperature OoC Detector temperature 410 C Fill Carrier gas Nitrogen 45 mvmin. Attention 5 x 104 Sample 0.3 L gas oil Percentage weights of n-paraffins with carbon number ranges C9-C14, C1S-C19 and C20-C32 were calculated. A computer programme was used to derive the statistical relationships between gas oils variables and their response towards pour point depressants and flow improvers.

2 318 INDIAN J. CHEM. TECHNOL., NOVEMBER 1995 G~ J G3 Gy G, Go III G2 G4 Variation SO ' G Test technique graphy Chroma to- G, III Table ~-Physico-chemical properties of gas oils method IP GJ Go G2 G4 Gq Gx 16 G, 7-4 'I Table G, Effect of additives on pour point and CFPP of gas oils Table 3-Selected variables Single variable Multiple variables Basic IBP 20% 90% FBP C,-C'4 C'" - CIY C2"--C,, Combined 20 vol. %-IBP 90 vol. %-20 vol. % FBB-YO vol. % FBP-20 \'01. % Two 90 vol. % & (FBP-90 vol. %) More FBP & 90 vol. % & (FBP-90 vol. %) YO vol. 'Yo& (FBP-20 vol. %) lbp & 90 vol. % & FBP 90 vol. %& FBP (90 vol. %-20 vol. %) & (FBP-20 vol. %) & (FBP-90 vol. %) FEP & (90 \'01. %-20 vol."!o) (90 vol. %-20 vol. %) & 90 vol. % & FBP IBP& FBP FBP & (FBP-YO Vul. %) YO Vol. % & (90 Vol. %-20 Vol. %) FBP & (FBP-20 Vol. %) (90 Vol. %-20 Vol. %) & (FBP-90 Vol. %) Results and Discussion The various gas oil distillation ranges evaluated, involves difference between recovery temperature for 90 vol % from that for 20 vol. % (90 vol. %-20 vol. %), difference between final boiling point from recovery temperature for 90 vol. % (FBP-90 vol. %) and final boiling points (FBP). Trials were conducted to derive statistical relationships' between gas oils distillation range as weil as chemical structure variables and their response towards pour point depressants and flow Improvers. I I I I 1 " nl' '1r I 'I "If 1111"'1 'l!~!l'j.!l

3 MOHAMED et al: LOW TEMPERATURE ADDITIVES 319 Derived regression analysis techniques Correlation via single variables-regression analysis, as a statistical technique, is generally used to examine results and draw conclusions about the functional relationships existing between the independent variable (e.g. gas oil distillation range and chemical structure variables) and the dependent response (e.g. gas oils response towards pour point depressants and flow improvers.) Results of correlations between gas oils response and selected variables, basic and combined (Table 3) via single linear regression analysis are listed in Table 4. It can be seen that the n-paraffin fraction within carbon number range C20-C32 is the only single variable which correlates well with the gas oils response to the two additives. A good correlation means that F> 60, T> 3.35 and R2> 0.94, where as F is a statistical parameter which is taken to assess the whole model significance. T is a statistical parameter which is taken to assess the significance of each partial relation (i.e., with each of the independent variables). Positive T value indicates direct proportionality between Y and X, while negative value shows inverse proportionality. And R2 is the multiple correlation coefficient (coefficient of determination). According to their table values at a confidence level higher than 99.5%, this variable (C20 C32) is inversely proportional to the gas oils response. The derived linear models for the accepted correlation between this variable and gas oil response towards the two pour point depressant and flow improvers are as follows Pour point depression = (C2o-Cd (F= , R 2 = and confidence % > 99.5) Cold filter plugging point depression = (C2o-Cd (F= '3, R2 = and confidence % > 99.5) Representation of the observed pour point depression and cold filter plugging point results against calculated ones according to that derived model (Fig. 1) is supported its acceptance. Correlations via multiple variables-different trials were conducted to develop statistical model via multiple linear regression analysis. Y=A+a1Xl +a2x2 +cfdxm I The multiple variables which' are considered to derive the anticipated correlations are listed in Table 3, while Table 5 includes the derived correlations which shows that the multiple variable = final boiling point (FBP) and difference in recovery temperature for 90 vol. % and for 20 vol. % (90 vol. %-20 vol. %) have been found in good correlation the two additives. with. gas oils response towards Derived models for this correlation are as follows: Conf.%* 0.Q L.C 0/0, %-IBP) C~-CI IDIDID6.39 FBP, CI5-CI~ C211-C3~ (20"' L.c ( (FBP- %), C 97.5 wt.% vol. C /0, 99.5 wt.% IBP, C C Table 4-Correlation betweencombined gas oils response C L.C.: Low Confidence ( < 90 vol. %) and single variation via single linear regression analysis : Inversely op. : Directly Proportionality proportional Single variables

4 320 INDIAN J. CHEM.1ECHNOL., NOVEMBER 1995 a:;..,., / q fl "u,.,1 25/,1 "< <v, I /" 15 u~,o 10~, 10 I Addilivt'(B) 5~/ 01, , 19-, " Obseorvf'd Obst'rvt'd Obseorvt'd,J 25l Add/livt' (A) ~ 25l Addilivt' (B) Fig. I-Observed and calculated pour point and cold filter plugging point depression results via single linear regression analysis Fig. 2-0bserved and calculated pour point and cold filter plugging point depression results via multiple linear regression analysis able Table Correlation R2 0.70Q F between gas Additive oils B A response and multiple variable via multiple regression analysis Fig. 2 represents observed pour point and cold filter plugging point depression results against results calculated according to the'derived model. It can be seen that prediction via this model has good accuracy. Pour point depres~on = (FBP) (90 vol. %-20 vol. %) (F= , R2 = and confidence % > 99.5) Cold filter plugging point depression = (FBP)+O.2861 (90 vol. %-20 vol. %) (F= , R2 = and confidence %> 99.5) Validation limits The application of the derived correlation is only valid within a certain distillation and chemical structure ranges, which are used for the derivation and validation of the correlations. The validation limits are as follows: Variable Validation range IBP,oC vol. %, C vol. %, C FBP,oC C20-C32, wt. % References 1 Fields I E, Symposium on polymers in lubricating oils, American Chemical Society, Division of Petroleum Chemistry, Reprints, 7, Nos. 4-B, B-61, B-71, Luri L, The role of flow improver technology in the egyp. tian oil industry environment, Technical Seminar, Cairo, 5-6 July SeIirn M, Abou El Naga H H & Abd El Azim W M, Evaluation of chemical structure for gas oil and its rekltion to cold flow propenies, paper presented _at the 3rd International SelJllnar on -Development in the Automotive Industry, Fuels and Lubricants, Research Centre, Misr Petroleum Company, Feb HolderG A & Winkler J, J Inst Pet, 51 (1965) Abou EI Naga H H & Mohamed M M, Trans Egypt Sac ChemEng, 14(1988)67. 6 Coley T R, Low temperature operability of diesels, SAE Technical Paper Series, No. 0596, Damin B & Faure A, /ljew additives for diesel fuels cloud point depressants, SAE Technical Paper Series No. 1527, 8 Brown G I, "Lehmann E W & Lewtas K, Evaluation of diesel fuel cold flow the next frontier, SAE Technical Paper Series No. 0031, I I, II I I I l1ii11iuii, "I'"''''.;;.

5 MOHAMED et al.: I.DW TEMPERATURE ADDITIVES Brown G I & Tack R 1;>,An additive solution to the prolr lem of wax settling in diesel fuels, SAE Technical Paper Series No. 1652, David J C, The development of diesel crystal modification technology and its impact upon refinery economics, paper presented at the 7th International Seminar on New Development in Engine and Industrial Oils, Fuels and Additives, Research Centre, Misr Petroleum Company, Cairo, 6-8 March Luri L, The cold filter plugging point (CFPP) and the use of middle'"distillate flow improvers to cut deeper in the crude barre~ paper presented at the 2nd International Seminar on Performance Testing and Development of Lubricants, Fuels and Chemicals, Research Centre, Misr Petroleum Company, Cairo, 2-5 April Abou El Naga H H, Abd El Azim W M & Mohamed M M, J Chern TechnolBiotechno~ 34A(1984) Bemad 0 & Richard W M, Statistics in research (The Iowe State University Press, USA), 1982.