Enterprise-wide optimization of integrated planning and scheduling for refinery-petrochemical complex with heuristic algorithm

Abstract

This paper focuses on the integrated problem of long-term planning and short-term scheduling in a large-scale refinery-petrochemical complex, and considers the overall manufacturing process from the upstream refinery to the downstream petrochemical site. Different time scales are incorporated from the planning and scheduling subproblems. At the end of each discrete time period, additional constraints are imposed to ensure material balance between different time scales. Discrete time representation is applied to the planning subproblem, while continuous time is applied to the scheduling of ethylene cracking and polymerization processes in the petrochemical site. An enterprise-wide mathematical model is formulated through mixed integer nonlinear programming. To solve the problem efficiently, a heuristic algorithm combined with a convolutional neural network (CNN), is proposed. Binary variables are used as the CNN input, leading to the integration of a data-driven approach and classical optimization by which a heuristic algorithm is established. The results do not only illustrate the detailed operations in a refinery and petrochemical complex under planning and scheduling, but also confirm the high efficiency of the proposed algorithm for solving large-scale problems.

Graphical abstract

Keywords

planning / scheduling / refinery-petrochemical / convolutional neural network / heuristic algorithm

References

[1] Sahinidis N, Grossmann I, Fornari R, Chathrathi M. Optimization model for long range planning in the chemical industry. Computers & Chemical Engineering, 1989, 13(9): 1049–1063

[2] Moro L, Zanin A, Pinto J. A planning model for refinery diesel production. Computers & Chemical Engineering, 1998, 22: S1039–S1042

[3] Pinto J M, Joly M, Moro L F L. Planning and scheduling models for refinery operations. Computers & Chemical Engineering, 2000, 24(9–10): 2259–2276

[4] Li W, Hui C W, Li A. Integrating CDU, FCC and product blending models into refinery planning. Computers & Chemical Engineering, 2005, 29(9): 2010–2028

[5] Alattas A M, Grossmann I E, Palou-Rivera I. Integration of nonlinear crude distillation unit models in refinery planning optimization. Industrial & Engineering Chemistry Research, 2011, 50(11): 6860–6870

[6] Alattas A M, Grossmann I E, Palou-Rivera I. Refinery production planning: multiperiod minlp with nonlinear CDU model. Industrial & Engineering Chemistry Research, 2012, 51(39): 12852–12861

[7] Li J, Misener R, Floudas C A. Continuous-time modeling and global optimization approach for scheduling of crude oil operations. AIChE Journal, 2012, 58(1): 205–226

[8] Zhang L, Yuan Z, Chen B. Refinery-wide planning operations under uncertainty via robust optimization approach coupled with global optimization. Computers & Chemical Engineering, 2021, 146: 107205

[9] Castro P M, Grossmann I E, Zhang Q. Expanding scope and computational challenges in process scheduling. Computers & Chemical Engineering, 2018, 114: 14–42

[10] Dogan M E, Grossmann I E. A decomposition method for the simultaneous planning and scheduling of single-stage continuous multiproduct plants. Industrial & Engineering Chemistry Research, 2006, 45(1): 299–315

[11] Erdirik-Dogan M, Grossmann I E. Simultaneous planning and scheduling of single-stage multi-product continuous plants with parallel lines. Computers & Chemical Engineering, 2008, 32(11): 2664–2683

[12] Li Z, Ierapetritou M G. Production planning and scheduling integration through augmented Lagrangian optimization. Computers & Chemical Engineering, 2010, 34(6): 996–1006

[13] Mouret S, Grossmann I E, Pestiaux P. A new Lagrangian decomposition approach applied to the integration of refinery planning and crude-oil scheduling. Computers & Chemical Engineering, 2011, 35(12): 2750–2766

[14] Li J, Xiao X, Boukouvala F, Floudas C A, Zhao B, Du G, Su X, Liu H. Data-driven mathematical modeling and global optimization framework for entire petrochemical planning operations. AIChE Journal, 2016, 62(9): 3020–3040

[15] Wang Z, Li Z, Feng Y, Rong G. Integrated short-term scheduling and production planning in an ethylene plant based on Lagrangian decomposition. Canadian Journal of Chemical Engineering, 2016, 94(9): 1723–1739

[16] Zhao H, Ierapetritou M G, Shah N K, Rong G. Integrated model of refining and petrochemical plant for enterprise-wide optimization. Computers & Chemical Engineering, 2017, 97: 194–207

[17] Ketabchi E, Mechleri E, Arellano-Garcia H. Increasing operational efficiency through the integration of an oil refinery and an ethylene production plant. Chemical Engineering Research & Design, 2019, 152: 85–94

[18] Uribe-Rodriguez A, Castro P M, Gonzalo G G, Chachuat B. Global optimization of large-scale MIQCQPs via cluster decomposition: application to short-term planning of an integrated refinery-petrochemical complex. Computers & Chemical Engineering, 2020, 140: 106883

[19] Yang H, Bernal D E, Franzoi R E, Engineer F G, Kwon K, Lee S, Grossmann I E. Integration of crude-oil scheduling and refinery planning by Lagrangean decomposition. Computers & Chemical Engineering, 2020, 138: 106812

[20] Zhang L, Yuan Z, Chen B. Adjustable robust optimization for the multi-period planning operations of an integrated refinery-petrochemical site under uncertainty. Computers & Chemical Engineering, 2022, 160: 107703

[21] BaoYPengY WuCLiZ, eds. Online Job Scheduling in Distributed Machine Learning Clusters. IEEE INFOCOM 2018–IEEE Conference on Computer Communications. IEEE, 2018: 495–503

[22] WangSLi DGengJ, eds. Geryon: Accelerating Distributed CNN Training by Network-Level Flow Scheduling. IEEE INFOCOM 2020–IEEE Conference on Computer Communications. IEEE, 2020: 1678–1687

[23] Guo P, Cheng W, Wang Y. Hybrid evolutionary algorithm with extreme machine learning fitness function evaluation for two-stage capacitated facility location problems. Expert Systems with Applications, 2017, 71: 57–68

[24] Aoun O, Sarhani M, El Afia A. Investigation of hidden markov model for the tuning of metaheuristics in airline scheduling problems. IFAC-PapersOnLine, 2016, 49(3): 347–352

[25] Aytug H, Bhattacharyya S, Koehler G J, Snowdon J L. A review of machine learning in scheduling. IEEE Transactions on Engineering Management, 1994, 41(2): 165–171

[26] Bagloee S A, Asadi M, Sarvi M, Patriksson M. A hybrid machine-learning and optimization method to solve bi-level problems. Expert Systems with Applications, 2018, 95: 142–152

[27] Bengio Y, Lodi A, Prouvost A. Machine learning for combinatorial optimization: a methodological tour d’horizon. European Journal of Operational Research, 2021, 290(2): 405–421

[28] Misra S, Roald L, Ng Y. Learning for constrained optimization: identifying optimal active constraint sets. INFORMS Journal on Computing, 2022, 34(1): 463–480

[29] Alvarez A M, Louveaux Q, Wehenkel L. A machine learning-based approximation of strong branching. INFORMS Journal on Computing, 2017, 29(1): 185–195

[30] Morabit M, Desaulniers G, Lodi A. Machine-learning-based column selection for column generation. Transportation Science, 2021, 55(4): 815–831

[31] Hou Y, Wu N, Zhou M, Li Z. Pareto-optimization for scheduling of crude oil operations in refinery via genetic algorithm. IEEE Transactions on Systems, Man, and Cybernetics. Systems, 2015, 47(3): 517–530

[32] Zhao Z, Liu S, Zhou M, Abusorrah A. Dual-objective mixed integer linear program and memetic algorithm for an industrial group scheduling problem. IEEE/CAA Journal of Automatica Sinica, 2020, 8(6): 1199–1209

[33] McCulloch W S, Pitts W. A logical calculus of the ideas immanent in nervous activity. Bulletin of Mathematical Biophysics, 1943, 5(4): 115–133

[34] Hubel D H, Wiesel T N. Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. Journal of Physiology, 1962, 160(1): 106–154

[35] FukushimaKMiyake S. Neocognitron: a self-organizing neural network model for a mechanism of visual pattern recognition. In: Competition and Cooperation in Neural Nets. Heidelberg: Springer, 1982, 267–285

[36] LeCun Y, Bottou L, Bengio Y, Haffner P. Gradient-based learning applied to document recognition. Proceedings of the IEEE, 1998, 86(11): 2278–2324

[37] Krizhevsky A, Sutskever I, Hinton G E. Imagenet classification with deep convolutional neural networks. Communications of the ACM, 2017, 60(6): 84–90

[38] ZeilerM DFergus R. Visualizing and Understanding Convolutional Networks. European Conference on Computer Vision. Cham: Springer, 2014, 818–833

[39] SzegedyCLiu WJiaYSermanetPReedS AnguelovDErhan DVanhouckeVRabinovichA. Going Deeper With Convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2015, 1–9

[40] SimonyanKZisserman A. Very deep convolutional networks for large-scale image recognition. arXiv preprint, 2014, 1409.1556

[41] HeKZhangX RenSSunJ. Deep Residual Learning for Image Recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016, 770–778

[42] Ieracitano C, Paviglianiti A, Campolo M, Hussain A, Pasero E, Morabito F C. A novel automatic classification system based on hybrid unsupervised and supervised machine learning for electrospun nanofibers. IEEE/CAA Journal of Automatica Sinica, 2020, 8(1): 64–76

[43] Muzahid A, Wan W, Sohel F, Wu L, Hou L. CurveNet: curvature-based multitask learning deep networks for 3D object recognition. IEEE/CAA Journal of Automatica Sinica, 2020, 8(6): 1177–1187

[44] Tawarmalani M, Sahinidis N V. A polyhedral branch-and-cut approach to global optimization. Mathematical Programming, 2005, 103(2): 225–249

[45] Grossmann I E, Viswanathan J, Vecchietti A, Raman R, Kalvelagen E. GAMS/DICOPT: a discrete continuous optimization package. GAMS Corporation Inc, 2002, 37: 55

[46] Gupta O K, Ravindran A. Branch and bound experiments in convex nonlinear integer programming. Management Science, 1985, 31(12): 1533–1546

[47] Bussieck M R, Pruessner A. Mixed-integer nonlinear programming. SIAG/OPT Newsletter. Views & News, 2003, 14(1): 19–22

[48] Duran M A, Grossmann I E. An outer-approximation algorithm for a class of mixed-integer nonlinear programs. Mathematical Programming, 1986, 36(3): 307–339

[49] Geoffrion A M. Generalized benders decomposition. Journal of Optimization Theory and Applications, 1972, 10(4): 237–260

Acknowledgements

The authors gratefully acknowledge the financial support from the National Key Research and Development Program of China (Grant No. 2018AAA0101602).

Nomenclature

Main units in refinery and petrochemical

Unit	Abbreviation
Refinery units
Delayed coking unit	DCU
Wax oil hydrotreating unit	WHU
Hydrocracking unit	HDT
Catalytic cracking unit	CCU
Coking naphtha hydrotreating unit	CNHT
Naphtha hydrotreating unit	NHT
Dry liquefied gas desulfurization unit	LDU
Liquefied gas separation unit	LSU
Dry gas separation unit A	DSU A
Continuous reformer	CRU
Aromatics complex	ARU
Naphtha n-isomer separation unit	NSU
Jet fuel hydrogenation unit	JHT
Diesel hydrocracking unit A	DHT A
Diesel hydrocracking unit B	DHT B
Gas fractionation unit	GFU
Alkylation unit	AKU
Hydrogen deep recovery device	HDRU
Hydrogen recovery unit	HRU
Dry gas separation unit B	DSU B
Catalytic gasoline hydrogenation unit	CHU
Jet fuel hydrogenation unit	JHT
Coke hydrogen production unit	CHP
Sulfur recovery unit	SRU
Petrochemical units
Gasoline hydrogenation unit	GHT
Butadiene extraction unit	BEU
MTBE/butene-1 unit	MTBE
Styrene unit	STU
Full density polyethylene unit	FDPE
Acrylonitrile butadiene styrene unit	ABS
Acetonitrile unit	ATU
methyl methacrylate unit	MMA
Acrylonitrile polymerization unit	APU
High density polyethylene unit	HDPE
Polypropylene unit	PP

Main products from refinery and petrochemical

Refinery product		Petrochemical product
92# gasoline	Pitch	C5	C9+
95# gasoline	PX	Propylene	Butadiene
98# gasoline	Petroleum coke	Styrene	MMA
Jet fuel	Butane	Acetonitrile	Acrylonitrile
Diesel	Sulfur	LLDPE	HDPE
Liquefied petroleum gas (LPG)	Ammonia	PP
Benzene	Fuel oil
Methylbenzene (PhMe)

Sets and indices

C	Set for crude distillation units
O	Set for crude oils
T	Set for time periods
N	Set for time plots in scheduling of ethylene cracking
J	Set for time plots in scheduling of polymerization process
S	Set for stream
R	Set for raw materials
U	Set for processing units
M	Set for operation modes
I	Set for intermediate stream tanks
P	Set for products
K	Set for cracking furnaces
Q	Set for polymerization reactors
L	Set for polymer products
G	Set for grades of polymer products

Parameters

X_{p,t}	Price of product p at time t
X_{q,l,g,t}	Price of grade g in polymer l from reactor q at time t
X_{c,o,t}	Price of crude oil o for CDU c at time t
X_{r,t}^{\mathrm{p}\mathrm{u}\mathrm{r}}	Price of raw material r at time t
V_{c,o}^{\mathrm{i}\mathrm{n}\mathrm{v}}	Inventory cost of crude oil o in CDU c
V_p^{\mathrm{i}\mathrm{n}\mathrm{v}}	Inventory cost of product p
V_{q,l,g}^{\mathrm{i}\mathrm{n}\mathrm{v}}	Inventory cost of grade g in polymer l from reactor q
V_{c,o}^{\mathrm{p}\mathrm{r}\mathrm{o}}	Process cost of crude oil o in CDU c
V_u^{\mathrm{o}\mathrm{p}}	Operation cost of unit u
V_k^{\mathrm{o}\mathrm{p}}	Operation cost of furnace k
V_q^{\mathrm{o}\mathrm{p}}	Operation cost of reactor q
V_k^{\mathrm{c}\mathrm{n}}	Clean up cost of furnace k
V_{i,ii,k}^{\mathrm{c}\mathrm{h}}	Changeover cost from intermediate i to ii in furnace k
V_{q,l,g,g^{\prime }}^{\mathrm{c}\mathrm{h}}	Changeover cost from grade g to g' in polymer l from reactor q
A_c^{\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum capacity of CDU c
A_c^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum capacity of CDU c
V_{c,o}^{\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum purchase amount of crude oil o to CDU c
V_{c,o}^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum purchase amount of crude oil o to CDU c
I_{c,o}^0	Initial inventory of crude oil o to CDU c
I_{c,o}^{\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum inventory of crude oil o to CDU c
I_{c,o}^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum inventory of crude oil o to CDU c
x_{c,o,s}^{\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum cut point for stream s of crude oil o from CDU c
x_{c,o,s}^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum cut point for stream s of crude oil o from CDU c
a_{r,u}	Constant inlet ratio of raw material r to unit u
A_u^{\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum capacity of unit u
A_u^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum capacity of unit u
x_{u,m,s}^{\mathrm{f}\mathrm{i}\mathrm{x}\mathrm{e}\mathrm{d}}	Constant yield of stream s in mode m in unit u
I_i^0	Initial inventory of intermediate i
H	Time length of each time period
F_{i,k,n,t}^{\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e},\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum flowrate of intermediate i to furnace k in slot n at time t
F_{i,k,n,t}^{\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e},\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum flowrate of intermediate i to furnace k in slot n at time t
T_{k,n,t}^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum operation time for furnace k in slot n at time t
a_{i,k,s}	Coefficient of yield of stream s from intermediate i in furnace k
b_{i,k,s}	Coefficient of yield of stream s from intermediate i in furnace k
c_{i,k,s}	Coefficient of yield of stream s from intermediate i in furnace k
x_{q,l,g}^{\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum yield of grade g of polymer l from reactor q
x_{q,l,g}^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum yield of grade g of polymer l from reactor q
T_{q,l,g,g^{\prime }}^{\mathrm{c}\mathrm{h}\mathrm{a},\mathrm{f}\mathrm{i}\mathrm{x}\mathrm{e}\mathrm{d}}	Constant changeover time from grade g to g' of polymer l in reactor q
A_q^{\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e},\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum capacity of reactor q
A_q^{\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e},\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum capacity of reactor q
d_{q.l,g,t}^{\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum demand of grade g of polymer l from reactor q at time t
d_{q,l,g,t}^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum demand of grade g of polymer l from reactor q at time t
d_{p,t}^{\mathrm{m}\mathrm{i}\mathrm{n}}	Minimum demand of product p at time t
d_{p,t}^{\mathrm{m}\mathrm{a}\mathrm{x}}	Maximum demand of product p at time t

Binary variables

y_{c,o,t}^{\mathrm{p}\mathrm{u}\mathrm{r}}	Purchase decision of crude oil o for CDU c at time t
y_{c,o,t}^{\mathrm{p}\mathrm{r}\mathrm{o}}	Process decision of crude oil o for CDU c at time t
y_{u,m,t}^{\mathrm{m}\mathrm{o}\mathrm{d}\mathrm{e}}	Assignment of operation mode m to unit u at time t
y_{i,k,n,t}^{\mathrm{a}\mathrm{s}\mathrm{s}\mathrm{i}}	Assignment of intermediate i to furnace k in plot n at time t
y_{i,ii,k,n,t}^{\mathrm{c}\mathrm{h}}	Assignment of changeover from intermediate i to ii for furnace k in plot n at time t
y_{k,n,t}^{\mathrm{c}\mathrm{n}}	Assignment of clean up operation to furnace k in plot n at time t
y_{q,l,g,j,t}^{\mathrm{a}\mathrm{s}\mathrm{s}\mathrm{i}}	Assignment of grade g to polymer l from reactor q in plot j at time t
y_{q,l,g,g^{\prime },j,t}^{\mathrm{c}\mathrm{h}}	Assignment of changeover from grade g to g' of polymer l from reactor q in plot j at time t

Continuous variables

z	Total profit as the objective function
F_{c,o,t}^{\mathrm{p}\mathrm{r}\mathrm{o}}	Process flowrate of crude oil o for CDU c at time t
F_{c,o,t}^{\mathrm{p}\mathrm{u}\mathrm{r}}	Purchase amount of crude oil o for CDU c at time t
I_{c,o,t}	Inventory amount of crude oil o for CDU c at time t
x_{c,o,s,t}	Cut point of stream s from crude oil o for CDU c at time t
x_{c,o,s,t}	Yield value of stream s from crude oil o for CDU c at time t
F_{c,s,t}^{\mathrm{o}\mathrm{u}\mathrm{t}}	Outlet flowrate of stream s from CDU c at time t
F_{c,s,u,t}	Outlet flowrate of stream s from CDU c to unit u at time t
F_{c,s,p,t}	Outlet flowrate of stream s from CDU c to product p at time t
F_{u,t}^{\mathrm{i}\mathrm{n}}	Inlet flowrate to unit u at time t
F_{i,u,t}	Outlet flowrate of intermediate i to unit u at time t
F_{r,u,t}	Outlet flowrate of raw material r to unit u at time t
F_{u^{\prime },s,u,t}	Outlet flowrate of stream s from unit u’ to unit u at time t
F_{u,m,t}^{\mathrm{i}\mathrm{n}}	Inlet flowrate to mode m in unit u at time t
F_{u,s,t}^{\mathrm{o}\mathrm{u}\mathrm{t}}	Outlet flowrate of stream s from unit u at time t
F_{u,s,p,t}	Outlet flowrate of stream s from unit u to product p at time t
F_{u,s,i,t}	Outlet flowrate of stream s from unit u to intermediate i at time t
I_{i,t}	Inventory amount of intermediate i at time t
F_{k,s,i,t}	Outlet flowrate of stream s from furnace k to intermediate i at time t
F_{i,k,t}	Outlet flowrate of intermediate i to furnace k at time t
F_{i,q,j,t}	Outlet flowrate of intermediate i to reactor q in slot j at time t
T_{k,n,t}^{\mathrm{s}\mathrm{t}}	Start time of slot n in furnace k at time t
T_{k,n,t}^{\mathrm{e}\mathrm{n}\mathrm{d}}	End time of slot n in furnace k at time t
T_{k,n,t}^{\mathrm{d}\mathrm{u}\mathrm{r}}	Duration time of slot n in furnace k at time t
T_{k,n,t}^{\mathrm{o}\mathrm{p}\mathrm{e}}	Total operation time for furnace k in slot n at time t
F_{i,k,n,t}^{\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}}	Outlet flowrate of intermediate i to furnace k in slot n at time t
x_{i,k,s,n,t}	Yield of stream s from intermediate i in furnace k in slot n at time t
F_{k,s,n,t}^{\mathrm{o}\mathrm{u}\mathrm{t}}	Outlet flowrate of stream s from furnace k in slot n at time t
T_{q,j,t}^{\mathrm{s}\mathrm{t}}	Start time of slot j in reactor q at time t
T_{q,j,t}^{\mathrm{e}\mathrm{n}\mathrm{d}}	End time of slot j in reactor q at time t
T_{q,j,t}^{\mathrm{d}\mathrm{u}\mathrm{r}}	Duration time of slot j in reactor q at time t
x_{q,l,g,j,t}	Yield of grade g in polymer l in reactor q in slot j at time t
T_{q,j,t}^{\mathrm{c}\mathrm{h}\mathrm{a}}	Changeover time of reactor q in slot j at time t
F_{q,j,t}^{\mathrm{i}\mathrm{n}}	Inlet flowrate to reactor q in slot j at time t
F_{r,q,j,t}	Inlet flowrate to reactor q from raw material r in slot j at time t
F_{q,l,g,j,t}^{\mathrm{o}\mathrm{u}\mathrm{t}}	Outlet flowrate of grade g in polymer l from reactor q in slot j at time t
F_{q,l,g,t}^{\mathrm{s}\mathrm{a}\mathrm{l}\mathrm{e}}	Sale amount of grade g in polymer l from reactor q at time t
I_{q,l,g,t}	Inventory amount of grade g in polymer l from reactor q at time t
F_{p,t}^{\mathrm{s}\mathrm{a}\mathrm{l}\mathrm{e}}	Sale amount of product p at time t
I_{p,t}	Inventory amount of product p at time t

RIGHTS & PERMISSIONS

2023 Higher Education Press