Data-driven minimization of pump operating and maintenance cost

Zhang, Zijun; He, Xiaofei; Kusiak, Andrew

doi:10.1016/j.engappai.2015.01.003

Cited by 19 publications

(11 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the 70s, pump scheduling have gained a special interest of the researchers and practitioners [2]. The majority of the pump scheduling literature addresses water pumps that are used in water distribution through water networks [2]- [8]. Other applications of pump scheduling include water treatment [9] and oil transportation [8].…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

An MDP Model for Integrated Operational Decision Making Under Uncertainty in a Gas and Oil Separation Plant

2021

View full text Add to dashboard Cite

Oil and gas supply chain (SC) plays key role in the global economy. Gas and oil separation plants (GOSP) belong to the upper stream of the oil and gas SC. In this article we study a GOSP that pumps oil to a stabilizing plant through a 600 KM pipeline. The purpose of this research is to develop an integrated mathematical model to support operational decision making regarding the optimal pumps scheduling and the dosage level of a drag reduction chemical, which is injected in the oil pipeline to stabilize the pressure and achieve higher oil flow rate. To the best of our knowledge, there is no MDP model in the literature that jointly considers pumps scheduling and oil flow control. A novel Markov decision process (MDP) model and two intuitive heuristic polices are proposed and simulated based on historical data. The heuristic policies are to operate the oil pumps in a cyclic weekly or biweekly patterns and to use a maximum likelihood rule to select the dosage level. Compared to the heuristic policies, our results demonstrated that MDP can lead to a substantial amount of savings in terms of the total system operating and maintenance costs. It also provides some practical insights on the interaction ways between the frequency of operating the pumps, the maintenance costs and the costs of a chemical that is used to control the pressure and flow rate of oil in a 600 KM oil pipeline.INDEX TERMS Decision making under uncertainty, gas and oil separation plant (GOSP), Markov decision Process (MDP), scheduling

show abstract

Section: Literature Reviewmentioning

confidence: 99%

“…In other applications, reference [8] studied the operations and maintenance scheduling of pumps used in a wastewater treatment process. The objective was to minimize the pumps' energy consumption and maintenance costs, while maintaining the desired hydraulic workload of the pumps.…”

Section: Literature Reviewmentioning

confidence: 99%

An MDP Model for Integrated Operational Decision Making Under Uncertainty in a Gas and Oil Separation Plant

2021

View full text Add to dashboard Cite

show abstract

“…The first step in numerical simulation is the modeling of three-dimensional entities. It is an important factor affecting the lattice division and the accuracy of the final research results whether the model is accurate or not [22][23][24][25][26][27]. The simulation range of the pumping station is 162.5 m for the upper reaches of the river and 37.5 m for the lower reaches of the river.…”

Section: Model Establishmentmentioning

confidence: 99%

Study on the Rectification of Forebay in Pumping Station

Song

Pang

Shi

et al. 2018

Mathematical Problems in Engineering

View full text Add to dashboard Cite

The forebay of pumping stations is an important hydraulic structure that connects the channel with the inlet channel. Actual test observations and theoretical studies have shown that poor precursors produce backflow, vortex, and water flow disturbances in the forebay water. In this paper, taking a lateral inlet pump station as an example, we study the nonmeasures and five rectification measures—“Y” type diversion pier, “T” shaped diversion pier, narrow bottom hole, high and wide bottom, and diversion wall—through adopting the method of numerical simulation and model test. For the numerical simulation, the corresponding three-dimensional model is established by UG solid modeling software, and then the computational fluid is simulated numerically with CFX. Based on the analysis and comparison of the results during the test of numerical simulation and model test, the stability of the rectification measures is considered after taking into consideration the results of the uniformity test of the velocity distribution of the surface layer, the bottom layer, and the front section of each scheme. The proposed scheme 3 (“T” diversion pier) is regarded as the pumping station flow control measures.

show abstract

“…Different from the CBM strategies for degradation processes with more explicit physical meaning, purely data-driven maintenance models have also gained much attention. For example, based on the collected data of pump speed and the junction chamber level, Zhang et.al [15] employed a neural network algorithm and a hierarchical particle swarm optimization algorithm to schedule the maintenance of pumps. Baptista et.al [16] integrated the ARMA model with datadriven techniques to predict fault events, and then make maintenance decisions for the aircraft engines.…”

Section: Introductionmentioning

confidence: 99%

Condition-based maintenance with imperfect inspections for continuous degradation processes

Hao

Yang

Bérenguer

2020

Applied Mathematical Modelling

View full text Add to dashboard Cite

A condition-based maintenance (CBM) strategy is now recognized as an efficient approach to perform maintenance at the best time before failures so as to save lifetime cycle cost. For continuous degradation processes, a significant source of variability lies in measurement errors caused by imperfect inspections, and this may lead to "false positive" or "false negative" observations, and consequently to inopportune maintenance decisions. To the best of our knowledge, researches on CBM optimization with imperfect inspections remain limited for continuous degradation processes, even though the subject is of practical interest for the implementation of a CBM policy. Imperfect inspections are indeed imperfect but still return interesting information on the system degradation level, and making them perfect can be expensive. Therefore, we analyze the economic performance of a maintenance policy with imperfect inspections, and compare it with the classical policy with perfect inspections to see which policy offers the best benefit in a given situation. Furthermore, a CBM policy with a two-stage inspection scheme is proposed to take benefit of mixing both perfect and imperfect inspections in the same maintenance policy. Through numerical experiments and a real case study, it is shown that the policy with imperfect inspections can be better than the classical one, and that the proposed policy with a two-stage inspection scheme always leads to the minimum long run maintenance cost rate.

show abstract

Data-driven minimization of pump operating and maintenance cost

Cited by 19 publications

References 34 publications

An MDP Model for Integrated Operational Decision Making Under Uncertainty in a Gas and Oil Separation Plant

An MDP Model for Integrated Operational Decision Making Under Uncertainty in a Gas and Oil Separation Plant

Study on the Rectification of Forebay in Pumping Station

Condition-based maintenance with imperfect inspections for continuous degradation processes

Contact Info

Product

Resources

About