A tool for assessing the energy demand and efficiency of machining systems: Energy benchmarking

Accepted Manuscript A tool for assessing the energy demand and efficiency of machining systems: Energy benchmarking

Wei Cai, Fei Liu, Jun Xie, Peiji Liu, Junbo Tuo PII:

S0360-5442(17)31214-8

DOI:

10.1016/j.energy.2017.07.039

Reference:

EGY 11227

To appear in:

Energy

Received Date:

12 May 2017

Revised Date:

29 June 2017

Accepted Date:

08 July 2017

Please cite this article as: Wei Cai, Fei Liu, Jun Xie, Peiji Liu, Junbo Tuo, A tool for assessing the energy demand and efficiency of machining systems: Energy benchmarking, Energy (2017), doi: 10.1016/j.energy.2017.07.039

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

A tool for assessing the energy demand and efficiency of machining systems: Energy benchmarking Wei Caia, Fei Liua,*, Jun Xieb, Peiji Liua, Junbo Tuoa a. State Key Laboratory of Mechanical Transmission, Chongqing University, Chongqing, 400030, China. b. College of mechanical engineering, Chongqing University of Technology, Chongqing, 400050, China.

Abstract Energy benchmarking has been recognized as an effective analytical methodology and management tool to improve energy efficiency and performance. Many approaches to energy benchmarking have been applied in various fields. Machining systems, which are widely distributed and consume large amounts of energy with low efficiency, possess considerable potential for reductions in energy consumption. However, current research regarding the use of energy benchmarking for machining systems is insufficient due to the complexity and variety of energy consumption processes used in these systems. This paper proposes the use of energy benchmarking to strengthen the evaluation of energy demand and achieve efficiency improvements for machining systems. First, it analyses drivers for energy benchmarking and their characteristics. Next, an energy benchmarking framework for machining systems is presented. Then the concepts of static, dynamic, single-objective, multi-objective, product-based, and processbased energy benchmarking are discussed from three different perspectives: the motion, object, and application level. This lays a theoretical foundation for further energy benchmarking research. Finally, methods for developing energy benchmarking are also addressed including the prediction method, statistical analysis, and expert decision. The application of these methods to a real machining plant allows an analysis of the practicability of potentially saving energy through benchmarking. Keywords：Energy benchmarking; Machining systems; Energy assessments; Energy demand; Energy efficiency; Energy consumption

* Corresponding author. State Key Laboratory of Mechanical Transmission, Chongqing University, Chongqing, 400030, China. E-mail address: [email protected] (W.Cai),[email protected] (F.Liu)

1

ACCEPTED MANUSCRIPT 1 Introduction The consistent increase in the global demand for energy, the failure to provide this quantity, and the conditions for developing new energy sources have resulted in a dramatic increase in the cost of energy over past decades [1-2]. This has made issues relating to energy supplies more urgent, especially for fossil fuels. Currently, the most effective measures to solve this problem are developing alternative energy sources [3-4], and improving energy efficiency [5]. Improving energy efficiency is the most direct and effective measure when compared with developing of alternative energy sources due to significant economic barriers. In 2012, the United States Energy Information Administration published the Energy Yearbook showing that industrial electricity consumption is responsible for 31% of total electricity consumption. Electricity consumption for manufacturing accounted for 90% of this use, and electricity consumption for machine tools or machining systems accounted for 75% of electricity use for manufacturing [6-7]. Machine tools, the basic energy consumption devices in machining systems, are numerous. For example, machining workshops in China involve the use of more than seven million machine tools—the highest in the world—resulting in an enormous energy consumption. Assuming that the average power consumption of each machine tool is 10 kilowatts, the total power consumed by these tools is approximately 70 million kilowatts, which is more than three times than the total capacity of the Three Gorges power station, the largest hydropower station in the world [8–9]. Besides, numerous surveys have shown that the energy efficiency of machining processes is very low, generally less than 30% [10-11]. Hence, reducing energy consumption and improving energy efficiency is an urgent problem for machining systems. Therefore, in this paper, we propose energy benchmarking as a method to assess the energy demand and efficiency in machining systems, contribute to strengthening energy management, and monitoring and improving energy efficiency. The framework of this paper is organized as follows (Fig. 1).

2

ACCEPTED MANUSCRIPT Energy benchmarking concepts Motion level Static energy benchmarking Dynamic energy benchmarking Object level Single-objective energy benchmarking Multi-objective energy benchmarking Application level Product-based energy benchmarking Process-based energy benchmarking

Framework Related concept of energy benchmarking Framework

Methods

Characteristics Complexity Multiple level Relevance

Prediction method Statistical Analysis Expert Decision

A

0-0.55

B

0.56-0.85

C

0.86-1.15

D

1.16-1.45

E

Over 1.45

Energy Benchmarking

Kwh >2.80 2.70 2.60 2.50 2.40

Application and analysis

2.30 2.20 2.10 2.00

Theoretical background

1.90 1.80

Energy measurement and monitoring Energy modelling and optimization Energy evaluation and energy-saving strategy Energy benchmarking Contributions

1.70 1.60 1.50

A Tool for Assessing the Energy Demand and Efficiency in Machining Systems

Fig. 1 Research Framework

2 Theoretical background Currently, massive methods for energy measurement, monitoring, modelling, and optimization have been applied to machining systems to improve their energy performance. These methods are useful but are not effective in measuring the energy consumption demand and for the application of specific constraints to energy use. Through a summary of previous methods, energy benchmarking is proposed to have particular advantages for analysing machining systems.

2.1 Energy measurement and monitoring Energy measurement and monitoring are important measures affording energy-related data support for machining systems. Machining environments incorporate a wide range of stakeholders who could benefit from energy-related data. For example: 

Senior management could benefit from the establishment environmental performance goals [12].



Energy managers could benefit from monitoring the implementation of goals to reduce energy consumption in machining processes.



Industrial designers could benefit from the energy efficient design of machining products.



Process managers could benefit from applying efficiency measures to production processes [13].



Machine tools operators could benefit from accessing relevant information for machining 3

ACCEPTED MANUSCRIPT process decisions. 

Suppliers could benefit from a better understanding of how to meet customer requirements [12]. For a machining plant, the systematic improvement of processes that exceed efficiency

measures requires the measurement of the energy consumption of machining systems or processes [14–15]. Monitoring is an extension of this measurement process, which involves additional record keeping, comparison, and visualization of these measurements [13,16]. Energy monitoring is necessary for activities that consume energy in machining systems. For example, Vijayaraghavan proposed an automated energy monitoring system for machine tools [17]. Lanz analysed the impact of energy measurements on machining operations [18]. Li introduced a multiscale statistical process for monitoring machining processes [19]. Behrendt developed an energy consumption monitoring procedure for machine tools [20]. Hu addressed an on-line approach for energy efficiency monitoring of machine tools [21]. Bornschlegl illustrated methods of energy measurement and an approach for the sustainable energy planning of manufacturing technologies [22]. Liu presented a novel approach for acquiring real-time energy efficiency measurements for machine tools [23]. Currently, studies on energy measurement and monitoring have had great success in laying a foundation for energy-efficient machining.

2.2 Energy modelling and optimization Energy modelling and optimization are regarded as basic measures for energy-efficient machining. Energy consumption is equal to power multiplied by time, while power is equal to force multiplied by speed. This force reflects the deformation of metal material and the speed reflects the variation of process parameters [7], which is the theoretical basis for building energy consumption models. Existing energy consumption models for machining can be roughly divided into three categories: linear type cutting energy consumption models based on the material removal rate (MRR); detailed parameter type cutting energy consumption correlation models; and process oriented machining energy consumption models [7]. These models can be summarized as shown in Table 1. Using these energy consumption models as a basis, extensive research has also been conducted for energy optimization models to reduce the energy demand. For example, Bi proposed 4

ACCEPTED MANUSCRIPT the optimization of machining processes from the perspective of energy consumption and provided an analysis of this case [24]. Wang addressed multi-objective optimization of machining parameters considering energy consumption [25]. He introduced an energy-responsive optimization method for machine tool selection and operation sequences in flexible machining job shops [26]. Feng integrated an energy, economy, and environmental analysis and an optimization of the energy supply system of a manufacturing plant [27]. He proposed Pareto fronts for machining parameters as a trade-off between energy consumption, cutting force and processing time [28]. Hu presented a method for minimizing the machining energy consumption of a machine tool by sequencing the features of a part [29]. Therefore, it is found that the development of energy modelling and optimization tools has been very rapid and they have caused concerns about the energy consumption of machining processes. Table. 1 Summary of existing energy consumption models for machining Types of energy consumption model

Scholars Gutowski et al. [30,31,32]

The linear type cutting energy consumption models based on MRR

Kara and Li [33] Li et al. [34] Diaz et al. [35]

Energy consumption model based on metal deformation theory

Detailed parameter type of cutting energy consumption correlation models

Energy consumption model based on the amount of tool wear

Munoz and Sheng [36] Kishawy et al. [37] Cuppini et al. [38] and Shao et al. [38]

Yoon et al [40,41] Energy consumption model based on the cutting force

Mohammed et al. [42]

Energy consumption model based on the main cutting parameters

Guo et al. [43]

Process oriented, machining energy consumption model

Liu and Liu [44] Lv et al. [45]

Overview

Models

Proposing a functional relationship between energy consumption and the material removal rate in machining processes Proposing a similar empirical model in which power is inversely proportional to the mass rate removal Proposing an improved specific energy consumption (SEC) model considering the spindle rotation speed for aircutting status

𝑃 = 𝑃0 + 𝑘 ∙ 𝑀𝑅𝑅 𝑃0 𝑆𝐸𝐶 = +𝑘 𝑀𝑅𝑅

Proposing a similar model Establishing an energy calculation model using the cutting force and the material removal rate vector Developing an energy model as a function of the volume fraction and material properties Proposing a face milling, cutting energy consumption model which has a linear relationship between the cutting power and the amount of tool wear Introducing tool wear into the building energy consumption model for milling machines and founding that the material removal power increases with the flank wear of the tool Establishing a SEC model for band sawing with a significant effect on the SEC between wear and degradation of the blade Pointing out that SEC is not only related to the cutting parameters but also related to component size Establishing a power model for main drive systems that consider that the mechanical and electrical main transmission system is the main body consuming energy Proposing that each functional module of the machine tool represents one

5

𝑆𝐸𝐶 = 𝐶0 + 𝑆𝐸𝐶 =

𝐶1 𝑀𝑅𝑅

𝑃 𝑛 1 = 𝑘0 + 𝑘1 + 𝑘2 𝑀𝑅𝑅 𝑀𝑅𝑅 𝑀𝑅𝑅 𝑘2 = 𝑃𝑠𝑡𝑎𝑛𝑑𝑏𝑦 + 𝑏

1 𝑆𝐸𝐶 = 𝑘 +𝑏 𝑀𝑅𝑅 𝐸𝑐𝑢𝑡 = 𝜏 ⋅ 𝑉𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 cos (𝛽 ‒ 𝛾)𝑐𝑜𝑠𝜂𝑠𝑐𝑜𝑠𝜆 + cos (𝜑 + 𝛽 ‒ 𝛾)𝑠𝑖𝑛𝜂𝑠𝑠𝑖𝑛𝜆 ( )× cos (𝜑 + 𝛽 ‒ 𝛾) 𝑠𝑖𝑛𝜑𝑐𝑜𝑠𝜆 𝐸𝑐𝑢𝑡 = 𝐸𝑃 + 𝐸𝑆 + 𝐸𝐷 𝐸𝑐𝑢𝑡 =

{

𝐷𝐾ℎ ‒ 𝑐[𝑐𝑜𝑠 𝜑 ‒ 𝑐𝑜𝑠 (𝜑𝑖𝑛 + 𝜓)] 2𝑎𝑒

+

}

𝐷𝜇𝐻𝑉𝐵𝜓 × 𝑉𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 2𝑎𝑒𝑓𝑧

𝐸 = 𝐸𝑐𝑜𝑛𝑠𝑡 + 𝐸𝑠𝑝𝑖𝑛𝑑𝑙𝑒 + 𝐸𝑓𝑒𝑒𝑑 + 𝐸𝑐𝑢𝑡

𝑆𝐸𝐶 =

𝑆𝐸𝐶 =

𝐶1

(𝑃0𝑡3 +

𝐶

𝑣𝑐 ∙ 𝑣𝑓 ∙ 𝑎𝑝 𝐸=

∫

∫

𝑡3 0

𝐹 𝐴𝑐ℎ𝑖𝑝 𝐶

𝐶

𝐶5

+ 𝐶0 ∙ 𝑣 𝑐2 ∙ 𝑣 𝑓3 ∙ 𝑎 𝑝4 ∙ 𝑑

𝑡1 0

𝑃𝑖𝑛(𝑡)𝑑𝑡 + 𝑃0𝑡2 +

𝑃𝛼[𝑃𝑐𝑢𝑡(𝑡)]𝑑𝑡 +

∫

𝑡3 0

𝑃𝑐𝑢𝑡(𝑡)𝑑𝑡)

𝑃 = 𝑃𝑆𝑂 + 𝑃𝐿 + 𝑃𝐶𝐶 + 𝑃𝐶𝐹𝑆 + 𝑃𝑠𝑝𝑖𝑛𝑑𝑙𝑒 + (𝑃𝑥 + 𝑃𝑦 + 𝑃𝑧) + 𝑃𝑇𝑆 + 𝑃𝑇𝐶 + 𝑃𝑐𝑢𝑡

ACCEPTED MANUSCRIPT basic action element

2.3 Energy evaluation and energy-saving strategy Energy evaluation of machining systems is rapidly expanding, however, it is difficult to evaluate the energy efficiency of these systems due to the complexity and variability of energy consumption processes. Despite this, current studies about energy efficiency in machining systems have made progress, which has provided valuable insights into methods for approaching the assessment of energy consumption in machining systems and selecting efficient process plans; finding potential energy efficiency improvements; and developing energy benchmarks. These studies usually focus on one of two types of evaluations including the holistic energy efficiency evaluation of machining systems or a unit energy efficiency analysis approach [46]. Holistic evaluations comprehensively assess the energy efficiency of machining systems. For example, Bernard proposed a principal component analysis method for measuring the energy use intensity of manufacturing industries [47]. Duflou addressed a processes and systems approach for analysing the energy and resource efficiency of manufacturing processes [48]. For a unit energy efficiency analysis, Wang presented an integrated method for assessing the energy efficiency of machining workshops; and Balogun introduced a specific energy based evaluation of machining efficiency [49]. Schudeleit and Züst proposed various methods for the evaluation of energy efficiency of machine tools [50]. Schudeleit presented a total energy efficiency index for machine tools [51]. Moreover, to reduce energy consumption and further improve the energy efficiency of machining systems, other energy-saving strategies have been presented, such as assessment and modelling, software-based optimization, control improvements, cutting improvement and hardware-based optimization and design for the environment. Actually, these energy-saving strategies depend on the basic technologies mentioned in sections 2.1, 2.2, and 2.3. For example, Diaz introduced strategies for the minimum energy operation of precision machining [52]. Aramcharoen discussed critical factors for the energy demand modelling of Computer Numerical Control (CNC) milling and the impact of toolpath strategies [53]. Yoon reviewed energy-saving strategies and technologies aimed towards developing greener machine tools [54].

2.4 Energy benchmarking Energy benchmarking is part of the much wider use of benchmarks as a tool for evaluating 6

ACCEPTED MANUSCRIPT energy demand and energy efficiency. The concept of the energy benchmarking was proposed last century and has been studied worldwide since, resulting in remarkable progress for industrial production [55–56]. Currently, a number of methods for developing energy benchmarks have been successfully applied to various industries, such as the petrochemical industry, steel and cement industry and coal mining industry as shown in Table 2. The study of energy benchmarking has aroused extensive interest in recent years [57–58]. Table. 2. A selected summary of existing energy benchmarking studies in various energy-intensive Industries from the literature Section Petrochemical industry Steel and cement industry Coal Mine Industry Pulp and paper industry Environmental Protection industry Agricultural and food industry Others

Benchmarking method

Specific research object

References sources

Coupled cluster method Mathematical modelling Strategic energy review — Mathematical modelling — Mathematical modelling Mathematical modelling Analysis Mathematical modelling Analysis k-means Comparative analysis Comparative analysis Statistical analysis Comparative analysis Mathematical modelling Mathematical modelling System modelling Statistical analysis Analysis

Molecular systems Industrial glass furnaces Petrochemical applications Oil and gas wells and cement slurries Iron and steel production Cement grinding Mineral Comminution Dump trucks in mines Copper and gold ores Coal production Production of paper and board Paper mill Kraft pulping mill Wastewater treatment plants (WWTP) WWTP Dutch industry Frozen food Paddy production — — Various industries in Taiwan

Řezáč J [59] Sardeshpande V [60] Rikhtegar F [61] Saleh F K [62] Worrell E [13] Zeng X [63] Nadolski S [64] Sahoo L K [65] Ballantyne G R [66] Wang N [67] Laurijssen J [68] Zhang Y [69] Mateos-Espejel E [70] Jonasson M [71] Krampe J [72] Phylipsen D [24] Prakash B [73] Chauhan N S [74] Ke J [6] Saygin D [12] Chan D Y L [11]

As mentioned previously, the wide distribution and a high consumption of low-efficiency energy mean that machining systems offer considerable energy-saving potential. However, few methods are available for developing energy benchmarks, which has resulted in the inadequacy of reduction of energy consumption and energy efficiency improvements for machining systems. Currently, Liu has proposed a method for dividing manufacturing products into a variety of general and individual specific products and has presented this strategy for development energy benchmarks for different product types [75]. Cai proposed a new concept called the fine energy consumption allowance (FECA) and a method for developing FECA for specific workpieces. The FECA may contribute to strengthening energy monitoring and management and improving energy efficiency in the mechanical manufacturing industry [76]. Zhou proposed an energy consumption model for establishing an energy-consumption allowance for specific workpieces in a machining system, but only introduced a modelling method and did not produce benchmarks [77]. Cai addressed an energy management approach for the mechanical manufacturing industry through 7

ACCEPTED MANUSCRIPT developing a multi-objective energy benchmark [78], and further developed a dynamic energy benchmark for mass production in machining systems for energy management and energyefficiency improvement [79]. Realistically, however, these studies are not systematic or comprehensive enough to reveal the nature of energy benchmarking in machining systems such as characteristic, methods, rules and applications for machining systems.

2.5 Contributions With regards to the analysis of energy measurement and monitoring, energy modelling and optimization, as well as energy evaluation and energy-saving strategies, these methods provide assistance for reducing energy consumption and improving energy efficiency. In this paper, we further proposed the application of energy benchmarking to strengthen the monitoring and evaluation of energy demand and improve energy efficiency in machining systems. This paper proposes systematic energy benchmarking using several concepts including an assessing of the characteristics of energy benchmarking, energy benchmarking frameworks, and the classification of energy approaches and methods for energy benchmarking, etc. Understanding the characteristics of energy benchmarking machining systems enables comprehension of the complex rules and features of this approach. Establishing energy benchmarking frameworks is important to provide a useful reference point for the further studying of energy benchmarking. The classification of energy benchmarking approaches and methods defines the different types knowledge about energy benchmarking and assists with developing energy benchmarks. The studies referred to in this paper provide a solid theoretical foundation for energy benchmarking research. Meanwhile, the application of this analysis of energy benchmarking not only has application to machining systems but may also play a role in developing benchmarking studies in other fields.

3 Characteristics of energy benchmarking for machining systems Characterizing the energy benchmarking is an indispensable step for constructing benchmarking frameworks, modelling, and determining methods for developing energy benchmarks for machining systems. Considering the energy consumption laws for a machining process, energy benchmarking of machining systems has complex, multi-level and correlative characteristics. 8

ACCEPTED MANUSCRIPT 3.1 The complexity of energy benchmarking Machined products comprise a number of components and parts; furthermore, the parts comprise a number of component parts; and by analogy, the last remaining part only consists of a number of components. Meanwhile, the energy consumption of the product to be machined comprises the energy consumption of the component, the part, assembly, storage, and transportation. Energy consumption of the component is composed of the energy consumption of the machining process and the auxiliary energy consumption for the air compressor, ventilation, air-conditioning, and illumination equipment and so on. Energy consumption of the part comprises the energy consumption of the component, the part, component assembly, assembly, storage, and transportation. Not only this, the constituent of energy consumption for the assembly, the storage, and the transportation is also complex and variable [75]. Fig. 2 shows the energy consumption constituents of a machined product and shows that the energy consumption for the machined product possesses a complex energy construction and numerous energy constituents resulting in difficulty in the development of energy benchmarks for machining systems. Product PREC

Component …

CEC

Component

Part

CEC

PEC

Part …

Part …

PEC

Component

Part

CEC

PEC

… CEC

Part

Part

…

PEC

…

Component

AEC

TEC and SEC

… PEC

…

Component

TEC and SEC SEC

…

…

Component

AEC

PEC

AEC

SEC

TEC and SEC

… CEC

CEC

SEC

PREC: Product Energy Consumption PEC: Part Energy Consumption CEC: Component Energy Consumption

AEC: Assembly Energy Consumption TEC: Transportation Energy Consumption SEC: Storage Energy Consumption

Fig. 2 Structure of a product and its energy consumption

Not only are the constituents of the machined product extremely complex and variable, but also the energy consumption processes. Taking the bottommost unit (a workpiece) of a machined product for example, the machining processes of this workpiece involves multiple machine tools, processes, and procedures. Even if the same workpiece is to be machined, the differences of energy consumption may be great because of the different machine tools and process plans used. Moreover, for just one machine tool has complex energy consumption processes. The machining 9

ACCEPTED MANUSCRIPT process of one machine tool could be divided into four types including standby, starting, idling, and cutting material processes as shown in Fig. 3. Therefore, energy benchmarking for machining systems is difficult.

Power demand (W)

Starting procedure

Cutting material procedure Idling procedure

Standby procedure

……

Idling procedure

Standby procedure

Time (s)

Fig. 3 A schematic diagram of a power profile for machining [76]

3.2 Multiple levels of energy benchmarking Multiple levels of energy benchmarking include multiple levels of objects and applications. From Fig. 4, machining products, in general, comprise a number of components and parts. In addition, the part comprises a number of components and parts, which indicates that the machined product, part, and component can respectively have energy benchmarks with different benchmark levels. Energy benchmarks for a component and a part are regarded as a subset of the energy benchmark for the machined product. In addition, energy benchmarking of a component is the lowest level of energy benchmarking, which is also known as the energy benchmark of a workpiece, which is an important basis for developing an energy benchmark for a part or a machined product. Defining an energy benchmark for a workpiece is the primary task for a machining plant or firm to begin an energy benchmarking study. Furthermore, for multiple levels of the application of energy benchmarks, taking the mechanical manufacturing industry in China, for example, involves a wide range of industries with vast production capacities, resulting in obvious differences in processing technologies, manufacturing equipment, and energy management systems for different regions and firms. Therefore, the energy consumption of machining processes for the each workpiece or machined product is obviously different, perhaps even hugely different between different regions and firms. Fig. 4 shows the capacity of the mechanical manufacturing industry in different regions with respect to processing technology, manufacturing equipment, and energy management systems. 10

ACCEPTED MANUSCRIPT Overall, provinces like JiangSu, ShangHai, GuangDong and ZheJiang possess advanced processing technologies and manufacturing equipment, as well as excellent energy management systems. In contrast, provinces like XinJiang, HaiNan, NingXia and Tibet possess backward processing technologies and manufacturing equipment, as well as substandard energy management systems. Therefore, differences in the energy consumption of machining processes for the same workpiece result in difficulties in implementing energy benchmarking for the same workpiece at the national level. To develop a reasonable and advanced energy benchmark for a workpiece, it is first necessary to study the energy benchmark for the plant or the firm, and then extend the benchmark to the industry or national level through science and technological developments.

HeiLongjiang JiLin XinJiang GanSu

QingHai

Inner Mongolia NingXia

BeiJing

HeBei Shan ShanDong Xi

HeNan ShaanXi

Tibet

LiaoNing

TianJin

S JiangSu ShangHai AnHui

SiChuan HuBei ChongQing Jiang ZheJiang HuNa Xi GuiZhou n FuJian Less-advanced YunNan TaiWan GuangXi  Processing technology GuangDong  Manufacturing equipment Hong Kong  Energy management system

More-advanced

HaiNan

South China Sea Islands

Fig. 4 Regional distribution of machining capacity

3.3 Relevance of energy benchmarking and production targets Conventional energy benchmarking is merely concerned with the energy consumption of production processes and whether energy consumption meets energy benchmarks. As we know, firms are not only concerned with energy consumption but also with other factors such as production costs, production efficiency, completion rates of products, product quality, and environmental performance. Generally, these production targets are closely related to process plans and the production requirements of firms. Production targets are different between different firms and there are large differences in these targets. Therefore, it is important for decisions about the problems experienced by firms to consider all these factors or objectives in terms of the 11

ACCEPTED MANUSCRIPT requirements of the entire firm. However, quantifying some of these objectives is difficult because they are multivariable. For example, environment performance involves ecological factors, the integrated use of natural resources, occupational health and safety and security, and these indicators are difficult to quantify. The completion rate of the product is closely related to temporal uncertainties (standby and idling time) and this uncertainty can be random and difficult to control. Therefore, energy benchmarks are closely related to production targets and this relationship is so complex that these production targets can be difficult to quantify.

4 Energy benchmarking frameworks for machining systems Energy benchmarking frameworks for machining systems are important to further energy benchmarking research and may reveal benchmarking characteristics through functional structures and modes of application to contribute to setting and applying benchmarks for energy management and energy efficiency improvements. This section includes a discussion of the concept of energy benchmarking frameworks.

4.1 The Related concept of energy benchmarking Energy benchmarking is a complicated engineering problem and development and application of energy benchmarking have obvious differences for different objectives. To construct a systematic energy benchmarking framework for machining systems, this paper proposes three levels of energy benchmarking based on benchmarking characteristics. These are namely the motion level, object level, and application level. 

Motion level energy benchmarks are used to describe the energy benchmarking characteristics where motion provides effective feedback for production requirements. This includes static energy and dynamic energy benchmarks.



Object level energy benchmarks are used to adapt different production targets to improve the management of energy consumption regarding different products. It comprises single-objective and multi-objective energy benchmarks.



Application level energy benchmarks consider production processes and energy consumption attributes that may effectively control energy consumption for the entire production process. These comprise product-based and process-based energy benchmarks. 12

ACCEPTED MANUSCRIPT 4.2 Energy benchmarking frameworks A summary of existing studies was used to propose an energy benchmarking framework including research objectives, application scope, research content and research goals, to lay a foundation for energy benchmarking research, as shown in Fig. 5. Research objectives

Workpiece or part

Application scope

Plant or firm

Industry

Country

Fundamental research

Level research

Method research

Characteristic study

Motion level

Product

 Characteristic of motion level  Static energy benchmarking  Dynamic energy benchmarking

Prediction method  Basis data acquirement  Prediction model  Development of energy benchmarking  …

Research content

 Complexity of energy benchmarking  Multiple levels of energy benchmarking  Relevance of energy benchmarking and production stargets

Object level  Characteristic of object level  Single-objective energy benchmarking  Multi-objective energy benchmarking

Statistical analysis  Energy data collection  Statistic models or methods  Development of energy benchmarking  …

Application level  Characteristic of application level  Product-based energy benchmarking  Process-based energy benchmarking

Research goals

Expert decision  Database establishment  Evaluation methods  Development of energy benchmarking  …

Energy Monitoring, management and energy efficiency improvement for machining systems

Fig. 5 An energy benchmarking framework for machining systems

It should be emphasized that (i) studying energy benchmarking and architecture from only one perspective makes it quite hard to account for the various the characteristics of energy benchmarking and the organic relationships that exist between these characteristics due to the complex, multi-level, and correlative characteristics of these benchmarks. Therefore, this paper proposes a consideration of energy benchmarking from three different perspectives: the motion level, object level, and application level to more fully describe and analyse energy benchmarks; (ii) Although motion, object, and application level energy benchmarks have obvious differences, they are closely related; and (iii) the proposed three energy benchmarking levels (motion, object, and application level) have a general structure, which is appropriate for various energy benchmarking contexts for machining systems.

5 Energy benchmarking for machining systems 5.1 Motion level: Static and dynamic energy benchmarking 13

ACCEPTED MANUSCRIPT At the motion level, energy benchmarking for machining systems includes static and dynamic energy benchmarking. Static energy benchmarking is a traditional approach that uses a simple numerical value to interpret the relationship between a product and its energy consumption, similar to the relationship between machining one workpiece and the energy consumption of that workpiece. Unnecessary energy consumption of the system can be identified by comparing the actual energy consumption with the energy benchmark. Energy consumption can be reduced by identifying useful measures using the static energy benchmarking to restrict excessive energy consumption in the machining process. Static energy benchmarking current is one of the most widely used approaches. Although static energy benchmarking plays a role in energy monitoring, energy management, and energy efficiency improvements. The effects of static benchmarking still have strong potential for further improvements to machining systems due to the complexity energy consumption processes. Dynamic energy benchmarking is a more advanced energy benchmarking approach and represents the continued development of static energy benchmarking. Dynamic energy benchmarking not only uses the function of static energy benchmarking, but also assesses the energy usage level of the machining process, by determining standard energy grades through an energy benchmarking rating system, along with the design of policies to further strengthen energy monitoring and management and energy efficiency improvements, as shown in Fig. 6 [79]. Dynamic energy benchmarking is based on the second rule of energy benchmarking, which was an important development for energy benchmarking of machining systems. More energy efficient

0-0.55 0.56-0.85 1.0 would be energy benchmarking of the product

0.86-1.15 1.16-1.45 Over 1.45 Less energy efficient

Fig. 6 A schematic diagram of a dynamic energy benchmarking [79]

5.2 Object level：Single-objective and multi-objective energy benchmarking Single-objective energy benchmarking is a conventional energy approach that only considers the energy consumption of the production process to achieve minimal energy consumption. 14

ACCEPTED MANUSCRIPT Realistically, besides low energy consumption for the production process, the mechanical manufacturing industry and machining systems also aim to lower production costs, increase production efficiencies, and improve environmental performance as much as possible. Therefore, the development of energy benchmarking should take into account a range of production targets. This gives the energy benchmarking the advantage of multiple objectives, as shown in Fig. 7. Generally, production targets are different across different firms or machining systems. Meanwhile, the quantification of some of these objectives is difficult. For example, environment performance involves ecological factors, integrated use of natural resources, occupational health and safety of operators and security. The completion rate of the product is closely related to temporal uncertainties such as standby time and idling time, and these uncertainties are often random and difficult to control. Therefore, an important issue is how to consider these production targets in multi-objective energy benchmarking and what decision-making methods should be used.

Fig. 7 Single-objective and multi-objective energy benchmarking

5.3 Application level：Product-based and process-based energy benchmarking At the application level, energy benchmarking comprises product-based and process-based energy approaches. Conventional energy benchmarking is associated with a product-based approach, which merely analyses the energy performance of a target product. Product-based energy benchmarking uses a simple numerical value to interpret the relationship between a product and its energy consumption, similar to the corresponding relationship between machining one workpiece and the energy consumption value for one workpiece machined [76]. For productbased energy benchmarking, unnecessary energy consumption for the whole production processes can be identified by comparing actual energy consumption with the benchmark and identifying useful measures for reducing unnecessary use. This indicates that product-based 15

ACCEPTED MANUSCRIPT energy benchmarks are only concerned with the total energy consumption. Product-based energy benchmarking have been widely applied in various industries, however, currently, there is still a lack of this type of energy benchmarking in the mechanical manufacturing industry or machining systems. Process-based energy benchmarking is obviously different to product-based benchmarking. It is more concerned with the benchmarking of each process and sub-process. Process-based energy benchmarking comprise one or more sets of energy data trees for the product rather than a simple energy value and may consider energy, time or process information for each process and sub-process. Furthermore, process-based energy benchmarks not only consider the total energy consumption of the workpiece but also highlights the energy consumption of each process and subprocess to achieve more effective energy management and energy efficiency improvements as shown in Fig. 8.

Fig. 8 Product-based and Process-based energy benchmarking

6 Methods for establishing energy benchmarking 6.1 Overview Methods for establishing energy benchmarking can be divided into prediction, statistical analysis, and expert decision methods. The prediction method focuses on building a mathematical model to comprehensively assessing energy consumption and is easier to establish a more operational energy benchmark focused on an objective. The statistical analysis method is used to acquire an energy benchmarking by analysing large amounts of energy consumption data related to workpieces. However, acquiring the energy consumption data is quite difficult, especially for the 16

ACCEPTED MANUSCRIPT new workpieces that are unprocessed in the machining plant. For the expert decision method, energy benchmarking is developed by analysing and assessing previous benchmarking and energy consumption data for other products or workpieces that have similar characteristics for each production process. However, this method requires strong professional capacity and it is difficult to find similar products (workpieces) and data to match with the benchmarking objective. Differences between the prediction, expert decision, and statistical analysis methods are shown in Table 3 and the application and comparison of these three methods are shown in Table 4. Table. 3. Differences among prediction method, expert decision, and statistical analysis Data Requirements

Applicability Methods

Prediction Method

New workpieces

Processed workpieces

√

√

√

√

√

Statistical Analysis Expert Decision

√

√

Low

High

√

Model Requirements Low

Medium

Reliability

High

Bad

Good

√

Excellent

√

√

√

√

√

Table. 4 Application and comparison of three methods Prediction method

Available

Step1

Step2.1

Step2.2

Step3

SSOPB energy benchmarking

●

●

●

○

○

SSOPB energy benchmarking

●

●

●

●

○

SMOPB energy benchmarking

○

○

○

○

○

DSOPB energy benchmarking

●

●

●

○

●

DSOPB energy benchmarking

●

●

●

●

●

DMOPB energy benchmarking

○

○

○

○

○

Expert Decision

Available

Step1

Step2

Step3

SSOPB energy benchmarking

●

●

●

○

SSOPB energy benchmarking

●

●

●

○

SMOPB energy benchmarking

○

○

○

○

DSOPB energy benchmarking

●

●

●

●

DSOPB energy benchmarking

●

●

●

●

DMOPB energy benchmarking

○

○

○

○

Statistical Analysis

Available

Step1

Step2

Step3

SSOPB energy benchmarking

●

●

●

○

SSOPB energy benchmarking

○

○

○

○

SMOPB energy benchmarking

●

●

●

○

DSOPB energy benchmarking

●

●

●

●

DSOPB energy benchmarking

○

○

○

○

DMOPB energy benchmarking

●

●

●

●

Note: Static, single-objective, product-based energy benchmarking (SSOPB energy benchmarking); Static, single-objective, processbased energy benchmarking (SSOPB energy benchmarking); Static, multi-objective, product-based energy benchmarking (SMOPB energy benchmarking); Dynamic, single-objective, product-based energy benchmarking (DSOPB energy benchmarking); Dynamic,

17

ACCEPTED MANUSCRIPT single-objective, process-based energy benchmarking (DSOPB energy benchmarking); Dynamic, multi-objective, product-based energy benchmarking (DMOPB energy benchmarking).

6.2 The prediction method Establishing an energy benchmarking for a product using the prediction method comprises three steps including data collection and analysis, modelling and determination of the energy benchmarking and development of the benchmarking rating system. (1) Data collection and analysis To conduct a problem analysis for the energy consumption of a machining system to develop an energy benchmarking, data collection and analysis are a basic first step. This includes data classification and data collection. Data classification distinguishes the energy consumption data and energy-related data(RED) to facilitate data collection. Energy consumption data has direct relevance to energy consumption in machining systems such as the power of the machine tool and power of the ancillary equipment. energy-unrelated data(RUD) is a kind of indirect energy consumption data such as the production time, numbers of machine tools, and the machining parameters. Data collection is the key to deciding the accuracy of the energy benchmarking established. In terms of actual conditions, some data may not be able to be obtained directly. However, at least some data is required to establish a data library or a function library such as the standby power or the load loss coefficient of the machine tool in advance. (2) Modelling and determination of the energy benchmarking Acquiring energy consumption data is difficult due to complexity and variability of machining processes. For a workpiece, the constitution of energy consumption is very complex. To model and determine the energy benchmarking, it is necessary to decompose the machining processes and energy consumption using a top-down method as shown in Fig. 9.

18

ACCEPTED MANUSCRIPT The whole machining process

Machining plant 1

1

2

...

O

1

...

1

Machining plant 2

2

...

f

Machining plant n

...

P

...

...

Total energy consumption

...

...

1

...

1

...

k

Energy consumption of plants

2

...

S

Energy consumption of machine tools

...

...

1

...

...

k

Energy consumption of basis process

Fig. 9. Decomposition of machining processes for energy consumption using a top-down method

The equations used to describe this base process in Fig. 9 is: (3)

𝐲 = 𝐹(𝐱) = 𝑈(𝐱) 𝐲(b) = 𝐹(b)(𝐱(b)) = 𝐹(b)(𝐺(𝐱)) = 𝑈(b)(𝐱) Where 𝐱 = (𝑥1,𝑥2,…,𝑥𝑚) and 𝐱

(b)

=(

(4)

(b) (b) 𝑥(b) 1 ,𝑥 2 ,…,𝑥 𝑚

)

are m-dimensional real vectors, which

influence the factors that represent the workpiece machining process (e.g. types of machine tools, number of machine tools machining parameters). 𝐺(𝐱) is the function translating the control variable 𝐱 to 𝐱

(𝑏)

(𝑏)

. 𝐲 = (𝑦,𝑦2,…,𝑦𝑛) and 𝐲

(𝑏) (𝑏) = (𝑦(𝑏) 1 ,𝑦 2 ,…,𝑦 𝑛 ) are n-dimensional real vectors that

represent the energy consumption and the corresponding energy benchmarking, respectively. 𝐲 = 𝐹(𝐱) and 𝐲(b) = 𝐹(𝑏)(𝐱(b)) are energy consumption models of machining systems and the energy (𝑏) (𝑏) benchmarking to be evaluated, respectively. 𝑈(𝐱) = 𝐹(𝐱) and 𝑈 (𝐱) = 𝐹 (𝐺(𝐱)) are the

composite function. (𝑏)

For the machining process, the energy consumption 𝐸 and energy benchmarking 𝐸

of the

base process is: 𝐸 = 𝐾(𝐲)

(5)

𝐸(𝑏) = 𝐾(𝑏)(𝑦(𝑏))

(6)

Where 𝐾(.)是 is the function of calculated energy consumption. Therefore, total energy consumption of the whole machining processes for the workpiece (𝑏) ETotal and energy benchmarking 𝐸𝑇𝑜𝑡𝑎𝑙 are: 𝑁

(7)

𝐸𝑇𝑜𝑡𝑎𝑙 = 𝐸1 + 𝐸2 + … + 𝐸𝑁 = ∑𝑖 𝐸𝑖 𝑁

(𝑏) = 𝐾(𝐸𝑇𝑜𝑡𝑎𝑙) = 𝐾(𝐸1 + 𝐸2 + … + 𝐸𝑁) = 𝐾(∑𝑖 𝐸𝑖) 𝐸𝑇𝑜𝑡𝑎𝑙

(8)

Step 2.2 Determination of process-based energy benchmarking For establishing a process-based energy benchmarking, which is actually a fine energy 19

ACCEPTED MANUSCRIPT benchmarking, it is necessary to consider, establish and determine the energy benchmarking of each machining process or sub-process. Hence, the energy benchmarking of each machining process or sub-process is: (9)

𝐲𝒊 = 𝑈𝑖(𝐱) (𝑏) 𝐲(𝑏) 𝒊 = 𝑈𝑖 (𝐱)

(10)

(𝑏) 𝐸(𝑏) 𝑖 = 𝐿(𝐲 𝒊 )

(11)

(𝑏) (𝑏) (𝑏) = 𝐸(𝑏) 𝐸𝑇𝑜𝑡𝑎𝑙 1 +𝐸 2 +…+𝐸𝑁

(12)

There are many methods for acquiring the energy consumption and benchmarking of the workpiece in machining systems. The above method only is to reveal the energy law from the perspective of systems engineering. A specific method can be illustrated in the case study. Step 3 Development of benchmarking rating system In terms of energy consumption laws proposed in Eq. 9, the benchmark rating system of workpieces in a machining system can be defined as a total of five grades from A to E (refer to Fig. 6 in Chongqing Machine Tool Works Co., Ltd, China). This classification method and grades for a benchmark rating system refer to energy labelling. As shown in Fig. 6, the value of benchmark rating (0-0.55, 0.560.85, 0.86-1.15, 1.16-1.45 and over 1.45) is not unique, and the value is determined from actual requirements of the firm, district government or national government. 𝜂𝐵𝑅 =

𝐸𝐴𝑐𝑡𝑢𝑎𝑙

(13)

𝐸𝐵

6.3 Statistical Analysis Statistical analysis mainly focuses energy consumption data that can be acquired or has been acquired to establish energy benchmarking, and this method includes three steps: (i) data collection, (ii) statistical analysis and determination of the energy benchmarking and (iii) development of a benchmarking rating system. Step 1 Data collection The energy consumption and non-energy consumption data of the machining system should be collected to perform the statistical analysis and further determine the energy benchmarking. These data are similar to the data for the prediction method, except more data, such as energy consumption of each machining process is added. There are various methods for acquiring this data such as establishing the energy monitoring and technical measurement devices. 20

ACCEPTED MANUSCRIPT Step 2 Statistical analysis and determination of the energy benchmarking To obtain an effective energy benchmarking, it is necessary to consider the actual energy demand. The specific method of statistical analysis is different for the single-objective and multiobjective energy benchmarking. For the single-objective energy benchmarking, all energy consumption data collected is used and the energy benchmarking is established by a measure of the central tendency. The measurement of the central tendency is an important statistical method, which is used to measure the characteristics of data distribution comprehensively [80–81]. The measure of central tendency includes an arithmetical mean, geometric mean, median, and mode. An arithmetic mean suits a situation that requires the data have a balanced distribution and strong differences among the variable values [82]. The application conditions of a geometric mean have some limitations, and it is usually used to average a data set of speed or rate [83]. Due to an imbalance in the energy consumption data of the workpieces and no multiplicative relationship existing between the energy consumption data of the workpieces and the explanatory variables, the arithmetic mean and geometric mean is not suitable to be used for measuring the central tendency of energy consumption data for these workpieces. The means could reflect the average level of data, but cannot represent the majority. The median cannot be influenced the maximum or minimum and is suitable to represent the central tendency of data set characteristics with large differences. The mode is the value that occurs most frequently in a data set or probability distribution is also a suitable method to describe the central tendency of the data set [39]. The percentage rank of a number is the percentage of numbers in its frequency distribution, which is the same or lower than this value [84], and this reflects the location of specific data in the data set. It was found that different measures of central tendency have different characteristics and their scope of application and representation are different. Therefore, according to statistical analysis energy benchmarking may be established for machining systems. In addition, a regression model may also solve these problems considering a range of integrated factors: 𝑘

𝐸 = 𝑐0 + 𝑐1𝑥1' + 𝑐2𝑥2' + … + 𝑐𝑘𝑥𝑘' + 𝜀 = 𝑐0 + ∑𝑖 = 1𝑐𝑖(

𝑥𝑖 ‒ 𝑥𝑖 𝑆𝑖

)+𝜀

(14)

Where E is the energy consumption; 𝑐0 is the intercept; 𝑐𝑖 is the standardized regression coefficient; 21

ACCEPTED MANUSCRIPT 𝑥𝑖(𝑖 = 1,2,…,𝑘) are the value of factors which affect E, such as manufacturing equipment, machining parameters, and production efficiency, etc.; 𝜀 is the random error. The energy benchmarking is: 𝑘

𝐸𝐷 = 𝐸𝐶 ‒ 𝑐1𝑥1' ‒ 𝑐2𝑥2' ‒ … ‒ 𝑐𝑘𝑥𝑘' + 𝜀 = 𝐸0 ‒ ∑𝑖 = 1𝑐𝑖(

𝑥𝑖 ‒ 𝑥𝑖 𝑆𝑖

)+𝜀

(15)

Where 𝐸𝐷 is the energy benchmarking; 𝐸𝐶 is the observed E; 𝑐𝑖 is a standardized regression coefficient; 𝑥𝑖(𝑖 = 1,2,…,𝑘) are the value of different factors which affect E, such as manufacturing equipment, machining parameters, and production efficiency, etc. For multi-objective energy benchmarking, on the basis of data acquisition, the energy benchmarking may be determined by various evaluation methods. In this paper, we will introduce the most common and popular entropy-weight and the fuzzy TOPSIS method. The entropy-weight and the fuzzy TOPSIS method includes (i) constructing an assessment matrix, (ii) finding 𝑦𝑖𝑗, (iii) calculating the entropy weight, (iv) acquiring a weighted fuzzy matrix, and (v) calculating the degree of relative similarity. Therefore, as decision objectives are different among different firms, the decision criteria can be determined in terms of actual requirements. The relative similarity degree 𝐶𝑖 with a value between 0 and 1 can be calculated by using Eq. (16). The closer this value gets to 1, the closer to the optimal level the assessed machining plan is. 𝐶𝑖 =

𝑠 ‒𝑖 𝑠 +𝑖 + 𝑠 ‒𝑖

(16)

𝑖 = 1,2,…,𝑚

The larger the value for 𝐶𝑖, the better the ith assessed machining plan is, and an optimal machining plan for energy consumption can be determined according to the MPi. Therefore, an energy benchmarking for a workpiece (or product) can be determined. Step 3 Development of benchmarking rating system Likewise, a benchmarking rating system may also be developed according to the method mentioned in section 6.2.

6.4 Expert Decision The prediction method and statistical analysis are the most popular methods used for setting benchmarking. However, the expert decision method is considered mediocre and the performance of this approach is usually not good. In this paper, the expert decision method may be described by three steps: (i) object matching, (ii) energy evaluation and determination of an energy benchmark, and (iii) development of a benchmarking rating system. Object matching is used to find and select a similar machining characteristic process (MCP) of 22

ACCEPTED MANUSCRIPT the manufactured product that possess corresponding energy consumption data. The MCP is the minimum machining unit such as the end of turning, the excircle of turning or hobbing. The selection of the MCP is important as if MCP values are more similar the benchmarking standard will be higher. Then, a comparison of the selected MCP and the MCP of the object is determined by experts and an estimation of energy consumption for the MCP of the object may be determined based on the energy consumption of the selected MCP. However, the error of this step is large. Therefore, the energy consumption of all MCPs and their energy benchmarking is also acquired. Besides, if necessary, a benchmarking rating system also may also be developed by the terms of section 6.2. The process of the expert decision method is as follows. 1

Object matching

Determination of similar MCPs

Decomposition of machining processes

MCPx1

MCPi

high low

low

MCPxj

high

similarity

MCPx2

The amount of MCPs:N

Selection of the similar MCPi

MCPx2

2

Energy evaluation of the MCPi

MCPi

similarity

MCPxj

Exj

Ei

Determination of the energy benchmarking

3

Determination of energy consumption for MCPi

Acquirement of energy consumption for all MCPi

Development of benchmarking rating system

Fig. 10 The process of the expert decision method

7 Application and analysis This section not only illustrates the establishment process for energy benchmarking but also demonstrates and analyses the practicability of the benchmarking for a real production process. Due to a wide variety of energy benchmarking processes and obvious differences in energy benchmarking between the firms, this section focuses on product-based energy benchmarking, dynamic energy benchmarking and their functions.

23

ACCEPTED MANUSCRIPT

Blank

Machining environment

Wire wheel

Fig. 11 The wire wheel used for the case study

Product-based energy benchmarking of a workpiece (a wire wheel) was established for the Chongqing Machine Tool Works Co., Ltd, China, using the prediction method to calculate energy consumption and determine a benchmarking. In this case, the wire wheel is as shown in Fig. 11, and the machining equipment is the CHK560CNC lathe. According to the prediction method, basic data needed to be collected and databases needed to be established beforehand. Basic data such as machining process parameters are shown in Table. 5. Table. 5 Machining process parameters for a wire wheel Content Step 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Spindle speed (rpm) Machining processes

Cutting times for machining

End of turning

Twice

Turning (Ø155 mm)

Thrice

Drilling

Once

Recessing

Four times

Turing hole (Ø30 mm) Recessing

Once Once

Turning (Ø155 mm)

Once

End of turning

Six times

Recessing

Four times

Recessing

Once

250 250 250 250 250 250 350 350 350 350 600 400 Exchanging plane 250 250 250 250 250 250 250 350 350 350 350 400

Feed (mm/r)

Depth of cut (mm)

0.22 0.22 0.23 0.23 0.23 0.15 0.04 0.04 0.20 0.22 0.20 0.22

2.0 2.5 5.0 5.0 1.0 – – – – – 6.0 –

0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.20 0.04 0.20 0.22 0.22

3.0 2.0 2.0 2.0 2.0 2.0 0.5 – – – – –

Meanwhile, according to characteristics of the machining process shown in Fig. 3, the machining processes of the wire wheel are decomposed into four types, namely standby, starting, idling and cutting material processes. Therefore, based on the collection of these machining parameters, it is necessary to establish basic energy consumption databases as shown in Table 6. According to the methods for establishing these databases, the standby power, starting energy 24

ACCEPTED MANUSCRIPT consumption, idling power and load loss coefficient for the machining of the wire wheel by the CHK560CNC lathe was acquired. Table 6 Methods for establishing basic energy consumption databases Databases Standby power

Starting energy consumption

Idling power

Load loss coefficient

Description A standby power database can be established by measuring the standby power of each machine tool before use. f(Mi)=Psbi Where Mi is the number of machine tool, Psbi is standby power of the machine tool The starting energy consumption is closely related to the spindle speed of the machine tool. When the spindle speed of the machine tool is determined, the starting energy consumption will be a constant, which means there is a functional relationship between energy consumption and spindle speed. Estj =g(nj) Where nj is spindle speed and Estj is the starting energy consumption. Similarly, the idling power is also closely related to the spindle speed of the machine tool. The idling power can be acquired by measuring the idling power of each level of speed of the machine tool. Pidj= h(nj) Where nj is spindle speed and j is the series of spindle speed. The load loss coefficient α of a machine tool is the ratio of additional load loss Pa to cutting power Pc [85]. The load loss coefficient is a constant between 0.15 and 0.25. Additional load loss is a linear or quadratic function of cutting power [75]. The load loss coefficient can be acquired by the methodology described in the literature [85].

The energy consumption of the whole of machining processes for the wire wheel can be determined using Eqs.(17)-(21) in terms of machining parameters and the basic database. 𝐸𝑊𝑖𝑟𝑒 𝑤ℎ𝑒𝑒𝑙 = ∑

𝑁𝑠𝑏 𝑁 𝑁 𝑁𝑐𝑚 𝐸 + ∑ 𝑠𝑡 𝐸𝑠𝑡𝑖 + ∑ 𝑖𝑑 𝐸𝑖𝑑𝑖 + ∑𝑖 = 1𝐸𝑐𝑚𝑖 𝑖 = 1 𝑠𝑏𝑖 𝑖=1 𝑖=1

(17)

𝐸𝑠𝑏 = 𝑓(𝑀𝑖) ∙ 𝑡𝑠𝑏

(18)

𝐸𝑠𝑡 = 𝑔(𝑛𝑖)

(19)

𝐸𝑖𝑑 = ℎ(𝑛𝑖) ∙ 𝑡𝑖𝑑

(20)

𝐸𝑐𝑚 = (ℎ(𝑛𝑖) + 𝛼𝑃𝑐) ∙ 𝑡𝑐𝑚

(21)

Where 𝐸𝑊𝑖𝑟𝑒 𝑤ℎ𝑒𝑒𝑙 is energy consumption of one wire wheel; 𝐸𝑠𝑏, 𝐸𝑠𝑡, 𝐸𝑖𝑑 and 𝐸𝑐𝑚 are the standby, starting, idling and cutting material energy consumption, respectively; 𝑁𝑠𝑏, 𝑁𝑠𝑡, 𝑁𝑖𝑑 and 𝑁𝑐𝑚 are the number of standby, starting, idling and cutting material processes, respectively. Therefore, product-based energy benchmarking of the wire wheel can be determined as 5 𝐸𝑊𝑖𝑟𝑒(𝑏) 𝑤ℎ𝑒𝑒𝑙 = 5.69 × 10 𝐽. Furthermore, for the dynamic energy benchmarking of the wire wheel,

using a benchmarking rating system can be established as shown in Fig. 12.

25

ACCEPTED MANUSCRIPT Benchmarking Rating System Firm：Chongqing Machine Tool Works Co., Ltd, China Workpiece：Wire wheel Energy consumption per unit gear More energy efficient

0-55 56-85 86-115 116-145 Over 145 Less energy efficient Energy consumption benchmark for the wire wheel ：0.15kWh Establishment time: 2016.12.02

First edition

Fig. 12. Benchmark rating system of a wire wheel in the Chongqing Machine Tool Works Co., Ltd, China.

The energy benchmarking plays a significant role in the energy assessment and improvement of energy efficiency. The operator can easily grasp the energy consumption requirements of the workpiece, and the benchmark rating and energy grades can be analysed via a comparison with the benchmark rating system to guide operators to avoid unreasonable operating parameters and select more reasonable machining plans and efficient process parameters. On the other hand, energy managers can master the overall energy consumption level of a workpiece and determine whether it meets eligibility criteria. The energy benchmarking and benchmark rating system also benefits the process of conducting energy audits, a collection of energy statistics, and energyefficiency analysis, aiding the decision-making processes of energy managers. The government can use benchmarks and the benchmark rating system to design relevant energy policies and standards for machining systems. For example, when the energy consumption of a machining system is below the energy grade for a benchmark rating system, the firm could be subject to financial and administrative penalties in accordance with the extent they breach the grade. Incentive schemes may be implemented for firms that satisfy energy benchmarks and grades. Additionally, benchmarking rating systems have a vital role in increasing the energy efficiency of machining systems. A machine tool spends large amounts of time in the standby and idling states because of poor energy-consciousness of operators, resulting in a significant waste of energy. Selected process parameters mainly depend on the consciousness and machining experience of the operator and satisfy the machining requirements but usually ignore energy-consumption issues. 26

ACCEPTED MANUSCRIPT Through the development of energy benchmarking, an operator can master achieving the energy consumption objectives and the grade compared with the benchmarking rating system efficiently, they can search for reasons for high energy consumption and subsequently adjust process parameters. In conclusion, energy benchmarking for a workpiece in a machining system is very important for achieving energy-efficient production.

8 Conclusions With a wide distribution and high amount of low-efficiency energy consumption, machining systems have considerable energy-saving potential. Currently, massive methods, such as energy measurement, monitoring, modelling and optimization, have been applied to machining systems to improve their performance. These methods are useful but are not effective in measuring energy consumption demand or applying specific constraints. Energy benchmarking has been recognized as an effective analytical methodology and management tool to improve the efficiency and performance of energy use. The study of energy benchmarking for machining systems is insufficient due to complexity and variety of energy consumption processes. This paper proposes the use of energy benchmarking to strengthen the evaluation of energy demand and energy efficiency of machining systems. First, this paper summarized some important methods such as energy measurement and monitoring, energy modelling and optimization, as well as energy evaluation and the development of energy-saving strategies for machining systems and analysed drivers for energy benchmarking in machining systems. Secondly, analysing three important characteristics of energy benchmarks in machining systems including complex, multi-level and correlative characteristics and how they contribute to constructing a framework, for modelling, and determining the methods for energy benchmarking in machining systems was completed. Then, an energy benchmarking framework for machining systems was developed for influencing energy benchmarking research. This revealed relevant energy benchmarking characteristic, functional structures, and application modes. Moreover, this paper discusses the concepts of the static, dynamic, single-objective, multiobjective, product-based and process-based energy benchmarking from the various perspectives of the motion, object, and application level This lays a theoretical foundation for energy benchmarking research. These contexts and applications of these benchmarking concepts were 27

ACCEPTED MANUSCRIPT illustrated in detail. Finally, to provide effective measures for energy benchmarking, this paper proposed three methods for developing energy benchmarking for machining systems including the prediction method, statistical analysis, and expert decision. Specific procedures of each method were described in detail and a comparison of these three methods was performed, which provided an important reference for energy benchmarking for machining systems. Furthermore, energy benchmarking was applied to Chongqing Machine Tool Works Co., Ltd., China, showing that this proposed method was feasible for establishing an energy benchmarking for a workpiece in a machining system and can play a crucial role in improving energy management and increasing energy efficiency. Future studies will focus on two aspects of energy benchmarking. First, an energy evaluation standard and certification process for machining systems will be considered using energy benchmarking processes. Second, data acquisition for establishing more basic databases will be continuously implemented.

Acknowledgments The authors acknowledge the technical support from the State Key Laboratory of Mechanical Transmission, Chongqing University. The project is supported by the National Natural Science Foundation of China (Grant No. 51375513), the National Hi-Tech Research and Development Program (863) (Grant No. 2014AA041506), and the Science and Technology Research Program of the Chongqing Municipal Education Commission (Grant No. KJ1709227).

References [1]

Velchev S, Kolev I, Ivanov K, et al. Empirical models for specific energy consumption and optimization of cutting parameters for minimizing energy consumption during turning. Journal of Cleaner Production 2014; 80:139-149. [2] Lee Y S, Tong L I. Forecasting nonlinear time series of energy consumption using a hybrid dynamic model. Applied Energy 2012; 94:251–256. [3] Yoon H S, Kim E S, Kim M S, et al. Towards greener machine tools–A review on energy saving strategies and technologies. Renewable and Sustainable Energy Reviews 2015;48:870–891. [4] Pao H T. Forecasting energy consumption in Taiwan using hybrid nonlinear models. Energy 2009; 34:1438–46. [5] Cagno E, Trianni A. Exploring drivers for energy efficiency within small- and medium-sized enterprises: First evidences from Italian manufacturing enterprises. Applied Energy 2013; 104: 276–285. [6] EIA, 2011. Annual Energy Review. http://www.eia.gov/totalenergy/data/annual/index.cfm. Last visited: March 2015. [7] Zhou L, Li J, Li F, et al. Energy consumption model and energy efficiency of machine tools: a comprehensive literature review. Journal of Cleaner Production 2016, 112: 3721-3734. [8] Andrea T, Enrico C, Ernst W. Empirical investigation of energy efficiency barriers in Italian manufacturing SMEs. Energy 2013;49: 444–458. [9] Liu F, Xu Z J, Dan B. Energy Characteristic of Mechanical Machining System and Its Application. China Machine Press; Beijing; 1995. (in Chinese) [10] Wang Q L, Liu F, Li C B. An integrated method for assessing the energy efficiency of machining workshop. 28

ACCEPTED MANUSCRIPT Journal of Cleaner Production 2013; 52:122–133. [11] Mouzon G, Yildirim M B, Twomey J. Operational methods for minimization of energy consumption of manufacturing equipment. International Journal of Production Research 2007;45(18-19):4247–4271. [12] Bogdanski G, Spiering T, Li W, et al. Energy monitoring in manufacturing companies – generating energy awareness through feedback, in: D.A. Dornfeld, B.S. Linke (Eds.), Leveraging Technology for a Sustainable World, Springer, Berlin/Heidelberg,2012, pp.539–544. [13] Spiering T, Kohlitz S, Sundmaeker H, et al. Energy efficiency benchmarking for injection moulding processes. Robotics and Computer-Integrated Manufacturing 2015, 36: 45-59. [14] J. Fresner. Energie-und Ressourceneffizienz in Unternehmen umsetzen, Nachhalt. Prod.2(2) (2011)58– 61. [15] Kara S, Bogdanski G, Li W. Electricity metering and monitoring in manufacturing systems. Glocalized solutions for sustainability in manufacturing 2011: 1-10. [16] Verein Deutscher Ingenieure VDI, Energiemanagement – Begriffe, No. 4602, Beuth Verlag, Düsseldorf, 2007. [17] Vijayaraghavan A, Dornfeld D. Automated energy monitoring of machine tools. CIRP AnnalsManufacturing Technology 2010, 59(1): 21-24. [18] Lanz M, Mani M, Leong S, et al. Impact of energy measurements in machining operations//ASME 2010 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. American Society of Mechanical Engineers, 2010: 867-873. [19] Li X, Yao X. Multi-scale statistical process monitoring in machining. IEEE Transactions on Industrial Electronics 2005, 52(3): 924-927. [20] Behrendt T, Zein A, Min S. Development of an energy consumption monitoring procedure for machine tools. CIRP Annals-Manufacturing Technology 2012, 61(1): 43-46. [21] Hu S, Liu F, He Y, et al. An on-line approach for energy efficiency monitoring of machine tools. Journal of cleaner production 2012, 27: 133-140. [22] Bornschlegl M, Bregulla M, Franke J. Methods-Energy Measurement – An approach for sustainable energy planning of manufacturing technologies. Journal of cleaner production 2016,135: 644–656. [23] Liu P, Liu F, Qiu H. A novel approach for acquiring the real-time energy efficiency of machine tools. Energy 2017,121:524–532. [24] Bi Z M, Wang L. Optimization of machining processes from the perspective of energy consumption: A case study. Journal of manufacturing systems 2012, 31(4): 420-428. [25] Wang Q, Liu F, Wang X. Multi-objective optimization of machining parameters considering energy consumption. The International Journal of Advanced Manufacturing Technology 2014, 71(5-8): 11331142. [26] He Y, Li Y, Wu T, et al. An energy-responsive optimization method for machine tool selection and operation sequence in flexible machining job shops. Journal of Cleaner Production, 2015, 87: 245-254. [27] Feng L, Mears L, Beaufort C, et al. Energy, economy, and environment analysis and optimization on manufacturing plant energy supply system. Energy Conversion and Management 2016, 117: 454-465. [28] He K, Tang R, Jin M. Pareto Fronts of Machining Parameters for Trade-off among Energy Consumption, Cutting Force and Processing Time. International Journal of Production Economics 2017,185:113-127. [29] Hu L, Peng C, Evans S, et al. Minimising the machining energy consumption of a machine tool by sequencing the features of a part. Energy 2017,121:292-305. [30] Gutowski, T., Dahmus, J., Thiriez, A., 2006. Electrical energy requirements for manufacturing processes. In: Proceedings of 13th CIRP International Conference on Life Cycle Engineering. Leuven, Belgium, pp. 5-11. [31] Gutowski, T. The Carbon and Energy Intensity of Manufacturing. 40th CIRP International Manufacturing Systems Seminar. Liverpool University, 2007. [32] Gutowski, T., Branham, M., Dahmus, J., Jones, A., Thiriez, A., 2009. Thermodynamic analysis of resources used in manufacturing processes. Environ. Sci. Technol. 43(5), 1584-1590. [33] Kara S, Li W. Unit process energy consumption models for material processes. CIRP Ann. Manuf. Technol 2011a; 61 (1): 37-40. [34] Li L, Yan J H, Xing Z W. Energy requirements evaluation of milling machines based on thermal equilibrium and empirical modeling. J Clean Prod 2013; 52: 113-121. [35] Diaz N, Redelsheimer E, Dornfeld D. Energy consumption characterization and reduction strategies for milling machine tool use. Glocalized Sol. Sustain. Manuf. Proc. CIRP Int. Conf. Life Cycle Eng. 2011, 263267. [36] Munoz A A, Sheng P. An analytical approach for determining the environmental impact of machining processes. J Mater Process Technol1995;53 (3-4):736-758. [37] Kishawy H A, Kannan S, Balazinski M. An energy based analytical force model for orthogonal cutting of metal matrix composites. CIRP Ann. Manuf. Technol 2004, 53 (1):91-94. [38] Cuppini D, D'errico G, Rutelli G. Tool wear monitoring based on cutting power measurement. Wear 1990,139 (2): 303-311. [39] Shao H, Wang H L, Zhao X M. A cutting power model for tool wear monitoring in milling. Int. Int. J. Mach. Tools Manuf. 2004;44 (14): 1503-1509. [40] Yoon H S, Moon J S, Pham M Q, et al. Control of machining parameters for energy and cost savings in micro-scale drilling of PCBs. J. Clean. Prod. 2013; 54:41-48. 29

ACCEPTED MANUSCRIPT [41] Yoon H S, Lee J Y, Kim M S, et al. Empirical power-consumption model for material removal in three-axis milling. J. Clean. Prod. 2014, 78: 54-62. [42] Mohammed S, Martin P, Hakan H, et al. Measurement of specific cutting energy for evaluating the efficiency of bandsawing different workpiece materials. Int. J. Mach. Tools Manuf. 2009. 49 (12-13), 958965. [43] Guo Y S, Loenders J, Duflou J, Lauwers B. Optimization of energy consumption and surface quality in finish turning. Procedia CIRP 2012. 1 (1): 512-517. [44] Shuang L iu, F Liu. Multi-period Energy Model of Electro-mechanical Main Driving System during the Service Process of Machine Tools. Journal of Mechanical Engineering, 2012, 21 -20. [45] Lv J X, Tang R Z, Jia S. Therblig-base d energy supply modeling of computer numerical control machine tools. J. Clean. Prod. 2014; 65: 168-177. [46] Wang Q, Liu F, Li C. An integrated method for assessing the energy efficiency of machining workshop. J. Clean. Prod. 2013, 52: 122-133. [47] Bernard J T, Côté B. The measurement of the energy intensity of manufacturing industries: a principal components analysis. Energy Policy 2005, 33(2): 221-233. [48] Duflou J R, Sutherland J W, Dornfeld D, et al. Towards energy and resource efficient manufacturing: A processes and systems approach. CIRP Annals-Manufacturing Technology 2012, 61(2): 587-609. [49] Balogun V A, Edem I F, Adekunle A A, et al. Specific energy based evaluation of machining efficiency. J. Clean. Prod.2016; 116: 187-197. [50] Schudeleit T, Züst S, Wegener K. Methods for evaluation of energy efficiency of machine tools. Energy 2015, 93: 1964-1970. [51] Schudeleit T, Züst S, Weiss L, et al. The Total Energy Efficiency Index for machine tools. Energy 2016; 102: 682-693. [52] Diaz N, Helu M, Jarvis A, et al. Strategies for minimum energy operation for precision machining. Laboratory for Manufacturing and Sustainability, 2009. [53] Aramcharoen A, Mativenga P T. Critical factors in energy demand modelling for CNC milling and impact of toolpath strategy. J. Clean. Prod. 2014; 78: 63-74. [54] Yoon H S, Kim E S, Kim M S, et al. Towards greener machine tools–A review on energy saving strategies and technologies. Renewable and Sustainable Energy Reviews 2015;48: 870-891. [55] Diego Iribarren, Mario Martín-Gamboa, Javier Dufour. Environmental benchmarking of wind farms according to their operational performance. Energy 2013; 61(1): 589–597. [56] Zhou X N, Liu F, Cai W. An energy-consumption model for establishing energy-consumption allowance of a workpiece in a machining system. J. Clean. Prod.2016; 135:1580–1590. [57] Yang T, Zhang X. Benchmarking the building energy consumption and solar energy trade-offs of residential neighborhoods on Chongming Eco-Island, China. Applied Energy 2016; 180: 792-799. [58] Chan D Y L, Huang C F, Lin W C, et al. Energy efficiency benchmarking of energy-intensive industries in Taiwan. Energy Conversion and Management 2014; 77: 216-220. [59] Řezáč J, Jurečka P, Riley K E, et al. Quantum chemical benchmark energy and geometry database for molecular clusters and complex molecular systems (www. begdb. com): a user’s manual and examples. Collection of Czechoslovak Chemical Communications 2008; 73(10): 1261-1270. [60] Sardeshpande V, Gaitonde U N, Banerjee R. Model based energy benchmarking for glass furnace. Energy Conversion and Management 2007; 48(10): 2718-2738. [61] Rikhtegar F. Energy benchmarking: a method incorporating energy optimization and its viability is exemplified in a petrochemical application. Chemical Engineering 2011; 118(12): 32-36. [62] Saleh F K, Teodoriu C. The mechanism of mixing and mixing energy for oil and gas wells cement slurries: A literature review and benchmarking of the findings. Journal of Natural Gas Science and Engineering, 2016. [63] Zeng X. Cement grinding technologies and energy benchmarking//Proceedings of the 2009 China Cement Grinding Technology Forum, Nanjng, Jiangsu, China, April 18th. < http://www. ccement. com/news/2009/4-18 C. 2009, 12177667. [64] Nadolski S, Klein B, Gong D, et al. Development and application of an energy benchmarking model for mineral comminution. SAG2015, Vancouver, 2015. [65] Sahoo L K, Bandyopadhyay S, Banerjee R. Benchmarking energy consumption for dump trucks in mines. Applied Energy 2014; 113: 1382-1396. [66] Ballantyne G R, Powell M S. Benchmarking comminution energy consumption for the processing of copper and gold ores. Minerals Engineering 2014; 65: 109-114. [67] Wang N, Wen Z, Liu M, et al. Constructing an energy efficiency benchmarking system for coal production. Applied Energy, 2016, 169: 301-308. [68] Laurijssen J, Faaij A, Worrell E. Benchmarking energy use in the paper industry: a benchmarking study on process unit level. Energy efficiency 2013; 6(1): 49-63. [69] Zhang Y, Hong M, Li J, et al. Benchmarking Analysis of Energy Efficiency Indicators in Paper Mill. BioResources 2016; 11(4): 9723-9740. [70] Mateos-Espejel E, Savulescu L, Maréchal F, et al. Base case process development for energy efficiency improvement, application to a Kraft pulping mill. Part II: benchmarking analysis. Chemical Engineering Research and Design 2011; 89(6): 729-741. [71] Jonasson M. Energy benchmark for wastewater treatment processes–a comparison between Sweden 30

ACCEPTED MANUSCRIPT and Austria. TEIE-5247, IEA, 2007. [72] Krampe J. Energy benchmarking of South Australian WWTPs. Water Science and Technology 2013; 67(9): 2059-2066. [73] Prakash B, Singh R P. Energy benchmarking of warehouses for frozen foods. Food Manufacturing Efficiency 2008, 1(3), 9. [74] Chauhan N S, Mohapatra P K J, Pandey K P. Improving energy productivity in paddy production through benchmarking—an application of data envelopment analysis. Energy Conversion and Management 2006; 47(9): 1063-1085. [75] Liu F, Zhou X N, Cai W. The complexity and strategy for establishing product energy consumption allowance in discrete manufacturing industry. J Mech Eng 2015; 10(19):138–45. (in Chinese) [76] Cai W, Liu F, Zhou XN, et al. Fine energy consumption allowance of workpieces in the mechanical manufacturing industry. Energy 2016; 114: 622–623. [77] Zhou X, Liu F, Cai W. An energy-consumption model for establishing energy-consumption allowance of a workpiece in a machining system. Journal of Cleaner Production 2016; 135: 1580-1590. [78] Cai W, Liu F, Xie J, et al. An energy management approach for the mechanical manufacturing industry through developing a multi-objective energy benchmark. Energy Conversion and Management, 2017, 132: 361-371. [79] Cai W, Liu F, Zhang H, et al. Development of dynamic energy benchmark for mass production in machining systems for energy management and energy-efficiency improvement. Applied Energy 2017; 202: 715-725. [80] Xiong J S, Wang J, Liu G. Statistics Tutorial. Zhejiang University Press. Hangzhou; 2007. [81] Joshua Kneifel, David Webb. Predicting energy performance of a net-zero energy building: A statistical approach. Applied Energy 2016;178: 468–483 [82] Tang, F., 2008. Reasonable application of arithmetic mean, mode and median. China Statistics 03, 49–50. (in Chinese) [83] Xin Y, Lu S, Zhu N, et al. Energy consumption quota of four and five star luxury hotel buildings in Hainan province, China. Energy and Buildings 2012; 45: 250-256. [84] Wikipedia, The Free Encyclopedia, 2010b. Cumulative Frequency Analysis. Retrieved on December 30, 2010 from . [85] Hu S H, Liu F, He Y. Peng B. Characteristics of additional load losses of spindle system of machine tools. Journal of Advanced Mechanical Design, Systems and Manufacturing 2010; 4 (7):1221–1233.

31

ACCEPTED MANUSCRIPT Highlights: Energy benchmarking for assessing the energy demand and energy was proposed. Drivers of energy benchmarking and their characteristics were analyzed. The energy benchmarking framework of machining systems was conducted. Some concepts about energy benchmarking (e.g. static energy benchmarking) were presented. Methods for developing the energy benchmarking were addressed.