Maintenance-centered Circular Manufacturing - Science Direct

Available online at www.sciencedirect.com

ScienceDirect Procedia CIRP 11 (2013) 23 – 31

2nd International Through-life Engineering Services Conference

Maintenance-centered Circular Manufacturing Shozo Takataa,* a

Department of Industrial and Management Systems Engineering, School of Creative Science and Engineering, Waseda University, Tokyo 169-8555, Japan

* Corresponding author. Tel.: +81-5286-3299; fax: +81-3202-2543. E-mail address: [email protected]

Abstract In the light of the growing problems of resource availability and environmental damage, it is indispensable to make best use of artifacts. For this purpose, maintenance and reuse are the most effective life cycle options. Maintenance and reuse are aimed at exhausting lives of artifacts, and use similar technologies such as condition diagnosis and restoration technologies. On the basis of this recognition, we reconsider the architecture of circular manufacturing and propose a concept of maintenance-centered circular manufacturing (MCCM), in which we do not discriminate between newly produced products and reused products as far as they can satisfy user requirements. As important issues in MCCM management, we discuss maintenance and reuse management from the aspect of operations and maintenance integration, and matching of the original use and the subsequent use. We also present maintenance and reuse management cases taking examples of the direct desulfurization facility at an oil refinery plant and the valve actuators of air-conditioning units.

© 2013 2013 The TheAuthors. Authors.Published PublishedbybyElsevier Elsevier B.V. Open access under CC BY-NC-ND license. © B.V. nd Selection of of thethe International Scientific Committee of the International Through-life Engineering Selectionand andpeer-review peer-reviewunder underresponsibility responsibility International Scientifi c Committee of"2 the “2nd International Through-life Services Conference" the Programme Chair – Ashutosh Tiwari. Engineering Servicesand Conference” and the Programme Chair – Ashutosh Tiwari Keywords: circular manufacturing; maintenance management; reuse management; operations and maintenance planning; life cycle simulation

1. Introduction Our daily lives are supported by various artifacts. We are able to perform day-to-day tasks through the functioning of various products and facilities, from those that support spaces such as homes, offices, shops, and factories to those that support networks such as power, water, transport, and communications systems. From the resource consumption perspective, a huge amount of resources is used to produce these artifacts. In light of the growing problems of resource availability and environmental damage, finding ways to reduce the environmental impact and resource consumption required to produce these artifacts has become an urgent issue for the manufacturing industry. To cope with this problem, the concept of circular manufacturing has been proposed, and various 3R (reduce, reuse, and recycling) technologies have been developed. Although maintenance has been recognized as an important life cycle option, which is a means to reduce environmental impact and resource consumption, discussions about maintenance and circular manufacturing have thus far

tended to be conducted separately. However, maintenance should be integrated as a central activity of circular manufacturing [1] and considered in the course of product life cycle planning [2], because the inner the loop is, the more effective circulation becomes in terms of energy and resource savings. In what follows, we propose the concept of maintenance-centered circular manufacturing (MCCM) in Section 2. Then, in Section 3, planning methods for maintenance and reuse are discussed. Finally, two case studies are presented in Section 4. 2. Concept of maintenance-centered circular manufacturing Since the 1990s, the concept of circular manufacturing has become increasingly important as a measure to reduce environmental impact and resource consumption in manufacturing. Figure 1 shows the concept of circular manufacturing. As depicted, there are multiple circulation loops such as maintenance, product reuse, part reuse, and

2212-8271 © 2013 The Authors. Published by Elsevier B.V. Open access under CC BY-NC-ND license.

Selection and peer-review under responsibility of the International Scientific Committee of the “2nd International Through-life Engineering Services Conference” and the Programme Chair – Ashutosh Tiwari doi:10.1016/j.procir.2013.07.066

24

Shozo Takata / Procedia CIRP 11 (2013) 23 – 31

materrial recycling.. To make efffective use of these options , we need to plan thee product liffe cycle app propriately. S Such ning activity is called life cycle c planning g [2]. In life ccycle plann plann ning, a holistic picture of th he product liffe cycle is draawn. hould contain product con ncepts, life cy ycle options, and It sh busin ness options. The product concept indicates functioonal, envirronmental, andd cost requireements and asssociated soluttions, inclu uding the valuee that the prod duct provides to the users. The life cycle c r enviro onmental load and options aare means to reduce resou urce consumpption, such ass maintenancee, upgrade, reeuse, r whhich correspon nd to various paths in circcular and recycling, manu ufacturing. Thhe business op ptions indicatee the delivery and billin ng methods ffor the produ ucts as well as the servvices integ grated into prooduct delivery.. Multiple M optionns should be implementeed in life ccycle plann ning. For exam mple, reused products p shou uld be recycleed at the end. e Accordiing to the concept of prroduct life ccycle plann ning, the impplementation of life cycle options mustt be discu ussed in an inntegrated way. However, th hey have thuss far been discussed inndependently. Maintenancee strategies hhave been selected withhout regard for f reuse, and d reuse has bbeen discu ussed without regard for reccycling processses. n reflection, the technolog gies adopted for maintenaance On upgraade, and reuse are similar.. We use such h technologiees as condition diagnossis, residual life estimation, disassem mbly, restorration (incluuding cleanin ng, adjustmeent, repair, and replaacement), insppection, and re-assembly. When prodducts continue to be usedd by the samee user, the actiivities to mainntain or en nhance the oriiginal function nality of the product p are caalled main ntenance and upgrade. However, H if the t products are collected and prrovided to other o users, these activvities consttitute reuse. W Whether the acctivities are caalled maintenaance or reuse, the samee technologies are necessary to exhaustt the item’’s life to the fuullest extent possible p throug gh restorationn and upgraade. On n the basis off this recognitiion, the archittecture of circcular manu ufacturing—thhe purpose of o which is to provide the requiired function to users whille circulating material—shoould

I the convenntional architeecture of circcular be reeconsidered. In manu ufacturing illu ustrated in Figgure 1, mainteenance, reuse, and recyccling are repreesented as suppplemental pro ocesses. Howeever, there is no reason n to discriminnate between newly produ uced produ ucts and reuseed products ass far as they satisfy user neeeds. In th his sense, we should integrrate manufactturing, reuse, and maintenance as ind dicated in Figgure 2. Here, only the shorrtfall p mo odules, partts and matterial (hereaafter, in products, collecctively called items) is new wly produced d, while itemss are circu ulated as much h as possible. W We call such a system MC CCM becau use the innerm most loop, that at is, maintenaance, is priorittized he most effiicient circulaation. To geet the maxim mum as th perfo ormance out of MCCM, w we need to manage it in n an integrated way. However, H succh an integraated managem ment d, although the method has nott yet beenn established manaagement meth hod for eachh life cycle option has been b discu ussed to a certaain extent. In n the next section, we discus uss the charactteristic featurees of maintenance and reuse r manageement and their integration n to realizze effective management off MCCM. nd reuse mannagement 3. Maintenance an M management m 3.1. Maintenance Operrations and ma aintenance Allthough the teechnologies us used for mainttenance and reeuse sharee similarities, maintenance management is different from f reusee managementt in that mainntenance should be executeed in parallel with op perations. A traditional framework of maintenance optim mization is rrepresented in n Figure 3. The figuree shows thatt an increasee in the degrree of preven ntive maintenance leads to decreased failure loss, but b it also lead ds to increased mainten nance cost. Thherefore, therre is an optim mum pointt in terms of degree d of prevventive mainteenance. Howeever, such a maintenan nce optimizatiion frameworrk does not give g much h attention to operations. T The aim of maaintenance sho ould be to o maximize th he value generrated by operaations rather than t m ccost. In this sense, s we sho ould to minimize the maintenance h consideratioon to the relattionships betw ween give more in-depth operaations and maintenance in discussiing maintenaance manaagement. Figure 4 showss the relationnships betweeen operations and maintenance using g production ffacilities as an n example. Th here are two t relationsh hips between operations and a maintenan ance:

Cost/Loss

pt of circular manufacturing Fig. 1 Connventional concep

tota al cost

maintenance co ost

failure loss

Degree of preventive maintena ance

Fig. 2 Concept of maintenance-ccentered circular manufacturing

o Fig. 3 Traditional framework oof maintenance optimization

25


sch heduling inteeractions and d interactions associated with faccility conditiions. From the operaation managgement perrspective, fulffilling the pro oduction requiirement is thee most imp portant constrraint. This som metimes causes conflict wiith the maaintenance reqquirement in terms t of sched duling. In thiss case, inssufficient maiintenance cou uld cause deeterioration oof the faccility conditioon to accelerate. This maay then resultt in a sho orter life, whhich leads to a higher life cycle cost ((LCC). Furrthermore, deeterioration of o the facility y conditions ccauses fun nctional debassement of thee facility, whiich could leaad to a deccreased produuction rate an nd/or lower yiield. By contrrast, if thee maintenancee plan is priorritized above operations, o w we may nott be able to ssecure enough h production capacity and suffer opp portunity losss because of the t interruptio on of productiion by maaintenance woorks.

Fig. F 4 Relationsh hips between operrations and mainttenance in the casse of productiion facilities

Pla anning proceddure To consider such relationships betweeen operationns and maaintenance in the maintenaance plan, wee need to inttegrate opeerations and m maintenance planning p as indicated in Figgure 5. The first step is to defin ne the objecctive functioon for opttimization. Thhis includes th he selection of o effects item ms and theeir weights as well as the setting s of consstraints. The eeffects item ms should bee listed to cover c all lossses associatedd with opeerations and maintenance.. They are id dentified by object typ pe based on the objectss affected by y events suuch as detterioration, ffailure, and maintenancee execution. Such objjects could bee humans, machines, m prod duced items, or the env vironment. Thhe effects item ms are also ch haracterized byy type of the effects such as safeety/quality, co ost, and timee. For exaample, in thee case of prroduction faccilities, if wee take pro oducts and tim me as the objeect and effect types, respecctively, thee effects item could be delaay in product delivery. How wever, e and escalators, itt could forr service faciliities such as elevators be the service iinterruption tiime. The weights are definned to sum m up various effects item ms, which are a quantifiedd with diffferent units ssuch as the maintenance m cost and the sservice plan inteerruption time. Operattion and maintenance m opttimization cann be achieved by minimizin ng the weighteed sum of effects items.. The constraiints, which haave to be takeen into acccount in the operation an nd maintenancce plan, shouuld be deffined also in thhis step. In the seconnd step, main ntenance plann ning and opeeration plaanning are cconducted in ndependently. While opeeration plaanning methodds differ depeending on the type of itemss to be maaintained and their operatio ons, there is a certain fram mework forr maintenancee planning, which w can bee generally aapplied reg gardless of thhe type of ittems. In maiintenance plaanning, critteria for treaatment appliccation, types of treatmentt, and tim ming and interrvals are deteermined for all deterioratioon and faillure modes thhat could occur ur in the items,, depending onn their me or chaaracteristics. The criteria for treatmentt could be tim con ndition, whicch correspon nd to time-b based mainteenance (TB BM) and conddition-based maintenance m (CBM), respecctively. The time criteriion is furtherr divided into o elapsed tim me and mount of operrations, such as the numbber of acccumulated am pro ocessed parts. The condition n criterion is also further ddivided into o symptoms and failures. If the criterion is failuree, it is usu ually called bbreakdown maintenance m (B BM). The typpes of

dure of integratedd operations and maintenance m plan n Fig. 5 Proced Table T 1 Types off treatment

o treatment Purpose of r restoration of de eterioration

Method d of treatment repair nt replacemen adjustment

m mitigation of detterioration facto ors cleaning, oiling reduction off operating stresss r removal of dete erioration factorss

provement material imp mprovement structural im

oses and meth hods of treaatment are cattegorized in teerms of purpo treaatment, as sh hown in Tablle 1. Timing g and intervaals are asssociated with application off treatment in n the case of TBM, and d execution off inspection orr monitoring in i the case of CBM. ming of treatments or inspections could be during d Tim opeerations, durin ng stoppage, and at the tim me of disasseembly. tween treatmeents or inspecctions. Intervals are thee periods betw m pplanning in this step is to id dentify The purpose of maintenance

26


the teechnically feaasible mainten nance policiess according too the charaacteristics of the deteriorattion and failu ure. For exam mple, TBM M is meaninggless for ran ndom failurees and CBM M is mean ningless for suudden failures. Th he third step is to integrate the operation n and maintenaance onsider variou us constraints and planss. In this step, we need to co tradeoffs between the operation plan and the maintenance pplan A already poointed out, wee need to conssider the scheddule [3]. As consttraints betweeen maintenan nce and operaations, as welll as ns of the item ms to their interrelationsships through the condition m Reegarding the former issuee, the questioon is be maintained. how to find the besst timing for executing e maintenance withh the minim mum effects oon operationss. A method of o identifyingg the ntenance opporrtunity windo ow is proposed d for this purppose main ue, we need to consider the [4]. Regarding thhe latter issu wing four cauusal relationsh hips: follow

Fig. 6 Time ssharing reuse

Functional level

100%

x

Require ed functional levvel of A

Requ uired functional level of B Use of user B

Use of user A

Fig. 7 Pass-on in the case that reeuse is enabled by the different functional reequirements 6

(a a)

Residual life

perations-induuced stress leaads to deteriorration of the • Op iteem’s conditionn. M treeatment restorres the producct’s condition.. • Maintenance • Deeterioration off the product’s condition leads to deebasement of ooperation quaality or operatiion rate. ondition also leads l to • Deeterioration off the item’s co acccelerated life consumption.

x

Use er A 4 2

User B tR

0

R manageement 3.2. Reuse Princciple of reuse Th he objective oof reuse is to exhaust e the liffetime of prodducts or parts p throughh multiple users u or mu ultiple produucts, respeectively, to redduce the numb ber of productts or parts neeeded to saatisfy the requuirements of a certain grou up of users ffor a certaiin period off time. Prod duct reuse is effective w when requiirements diffeer in terms of timing or am mount of usagee, as well as quality or functionalityy of the produ ucts. Part reusse is effective when thhere are differrences in thee lifetimes off the parts that constitutte the productss.

Replacem ment of user B

tL

6

(b b)

Residual life

we need to seelect With W regard to the first two relationships, r propeer operating conditions to o minimize the t stress andd to apply y proper treatm ment to maxim mize the effects of maintenaance. With h regard to thhe latter two relationships, r we need to deal with the tradeoff bbetween the lo osses induced d by deteriorattion, such as higher LC CC and decreaase of yield an nd operating rrate, and losses l inducedd by the execcution of main ntenance, suchh as lowerr operating raate. There are various strateegies to cope w with thesee problems. Inn the third steep, we need to o combine prooper strateegies and creaate feasible maanagement sceenarios. In n the fourth sttep, the scenar arios listed in the third stepp are evalu uated to determ mine whetherr the results saatisfy constraaints, and the best onee is selected.. If no scenaario satisfies the consttraints, the proocedure return ns to the secon nd or third stepp. In n the evaluatiion, we need d to assess all effects ittems selected in the firstt step by conssidering events such as failuures, generration of infe ferior productts, and servicce outages. IIt is difficcult to deal with all th hese factors using a sim mple math hematical moodel. Life cyycle simulatiion (LCS) iis a poweerful tool for solving such a problem an nd is discusseed in Section 3.3.

er A Use 4 User B 2 0

tE Exchange betw ween user A and d user B

tL

Fig. 8 Pass-on in the case that reuse is eenabled by the different the usage rates

orized into twoo types based on the princiiples Reeuse is catego used to achieving its purpose: the time-sharring type and d the me sharing iis usually ap pplicable onlyy to pass--on type. Tim produ uct reuse (with some exceeptions such as optional parts p needeed in specificc circumstancces; e.g., win nter tires). In this type of reuse, a pro oduct is shareed by multiplee users becausse of As shown in Figure 6, useer A their differing usaage periods. A u B use a washing mach chine in differrent time periods. and user For example, e user A uses a waashing machin ne in the daytiime, whereas user B uses it only in th the evening. In n such a case,, the total number of machines m is reeduced to on ne (from two)) by ng one machin ne shared betw ween users A and a B. havin Ho owever, in pass-on reuse, th the residual liffe of a producct or part is i subsequently used by anoother user or in another prod duct. In th he case of product reusse, this is realized thro ough differrences in the required funcctionality or amounts a of ussage amon ng users. In th he case of partt reuse, this iss realized thro ough the differing lifetim mes of the prooduct and parts. An n example off pass-on reusee is shown in n Figure 7, wh here reusee is enabled by the diffeerence in req quirements at the functtional level. The T figure inddicates functiional debasem ment versu us usage time. Here, we asssume, for simplification, that functtional debasem ment is propor ortional to tim me. The functio onal level required by user A is asssumed to be higher than that

27


Factors to be considered in reuse implementation Although reuse is an effective life cycle option, it is not easy to implement because of various conditions that need to be considered. Umeda et al. discussed the conditions for reuse applications using the following four factors: residual life, residual value, costs, and the supply and demand balance of reusable items [5]. Regarding lifetime, the residual life of the product or part needs to be longer than the guaranteed life for the next use. Regarding value or quality, the remaining value or quality of the item to be reused needs to satisfy the requirements for the next use. To implement reuse from an economic perspective, the cost of reuse has to be less than the sum of the cost of producing new products or parts and that of recycling. The supply and demand balance is the most critical factor to be considered. To realize reuse, we need to match the original use with the subsequent reuse in terms of the abovementioned three factors. For example, in the case of the reuse shown in Figure 7, we need to find a pair of users, A and B, which match each other. However, the functional requirements for the reused products dispersed depend on the users. Product return is also at the users’ discretion once the products have been sold. Therefore, matching the returned items with the demand for reused items is a challenge for the successful implementation of reuse. Reuse management policies As stated above, products are usually collected at users’ convenience when they stop using them. We call this case spontaneous collection. In this case, product providers do not have any control over the volume and quality of the returned products. This makes it difficult to match the original use with the subsequent reuse. To solve this problem, we propose planned circulation (also called forced circulation). In the case of spontaneous collection, products are usually owned by users, whereas in the case of planned circulation, they should be owned by the product providers. In the latter case, providers are able to control product collection because they have the authority to manage the products. Planned circulation can facilitate both the types of reuse explained in Figure 7 and

Figure 8. We applied this concept to copier photoconductor drum units [6] and to a group of copiers used in an office [7]. Another strategy to mitigate the difficulty in matching the demanded functionalities with those of the returned items is to increase the variety of supplies of the reused items in terms of functionalities. For this purpose, we propose the concept of module reconfiguration in remanufacturing, where the modules extracted from the returned products of different product generations are reconfigured to satisfy various user requirements [8]. 3.3. Integration of maintenance and reuse management to realize maintenance-centered circular manufacturing As stated already, maintenance and reuse are executed to exhaust the lives of items as much as possible. There is no fundamental difference between maintenance and reuse in terms of restoration technologies. If a single user keeps using items, or in case items are not portable (e.g., buildings), the activity is usually called maintenance; otherwise, it is called reuse. The applicable maintenance and reuse policies depend on the degradation characteristics of the items. When the state of the item is represented by functional levels and physical conditions, the degradation characteristics are divided into three types, as indicated in Figure 9. Type 1 is the case where physical deterioration leads to functional debasement, as in the case of cutting tools. Type 2 is the case where only physical deterioration occurs without functional debasement, as in the case of ink cartridges. Type 3 is the case where relative functional debasement is induced by the appearance of new initial state of the item

physical state

required by user B. When the functional level falls below the level required by user A, user A stops using the product. At this point, if the functional level is higher than that required by user B, the product can be reused by user B. Another example of pass-on reuse is shown in Figure 8, where reuse is enabled by the users’ differing amount of use per unit time. If user B uses the product at a higher rate than user A, and both user A and user B need the product until tL, we need three products to satisfy the needs of both users if we provide products to the users separately because user B exhausts the physical life of the product at tR. However, if we exchange the products between user A and B at tE, we only need two products instead of three. In both examples, we can increase the utilization rate of the product lifetime and, consequently, reduce the required number of products. The principles of reuse explained above can also be applied to part reuse, except that the lives of parts are exhausted by multiple products rather than multiple users.

type1: physical deterioration with functional debasement

initial state of the new model

type3: relative functional debasement owing to a new model type2: physical deterioration

functional level Fig. 9 Categorization of degradation characteristics Table 2 Applicable maintenance and reuse policies depending on degradation characteristics Degradation characteristics Type 1

Type 2

Type 3

Maintenance

Reuse

restorable

long use with restoration

time sharing or pass-on reuse to users with different usage rates

not restorable

(expendables)

cascade reuse

restorable

long use with restoration


not restorable

(expendables)


upgradable

long use with upgrade

time sharing or pass-on reuse with upgrade

not upgradable

not applicable

cascade reuse

28


models or new technology, but there is little physical deterioration, as in the case of microprocessors. Each type is further divided into two cases: restorable and not restorable. Applicable maintenance and reuse policies are indicated for each type of degradation characteristic in Table 2. In this case, reuse is differentiated from maintenance when the items are used by multiple users (in case of product reuse) or in multiple products (in case of part reuse). Although, as we have stated several times, there is no fundamental difference between maintenance and reuse from the aspect of restoration technologies, management technologies have been developed for the different aspects of maintenance and reuse. We have mainly discussed maintenance management in terms of how to restore items effectively and efficiently using criteria for treatment application, types of treatment, and the timing and intervals of treatment or inspection. However, reuse management has been discussed from the perspective of logistics such as timing and methods for product return and where the items are reused. To realize the concept of MCCM depicted in Figure 2, we need to integrate these technologies in a way that addresses the following questions. • Which item should be maintained and which should be replaced and reused according to the degradation characteristics of the items? • When should we stop using the items and pass them on for reuse depending on the states of deterioration or functional debasement of the items? • How we can effectively make use of reused items for maintenance? Development of technologies to deal with these issues will be a challenge for the future. 3.4. Life cycle simulation to evaluate circular manufacturing Circular manufacturing, by definition, involves closed loops in the flows of items. These loops are controlled by different maintenance and reuse policies depending on the item characteristics. It also involves various stochastic events such as failures and product returns. As a result, circular manufacturing behavior becomes complex and hard to evaluate. Recently, LCS has been developed as a powerful tool for solving this problem [3]. It can evaluate the dynamic behavior of circular manufacturing based on discrete event simulation. It simulates the flow of items through the various processes constituting circular manufacturing, such as usage, maintenance, product return, diagnosis, restoration, and production of items. By combining LCC and LCA (Life Cycle Assessment) data, LCS can evaluate the cost and environmental load that must be paid in exchange for providing items to a certain market for a certain period. To evaluate effectiveness of the maintenance policies, the Monte Carlo simulation technique is used to assess the effects of stochastically occurring events such as deterioration and failures [10].

4. Application examples Although there has not yet been a good example of MCCM, we would like to present two examples of maintenance management. One is an example of operation and maintenance integration applied to a direct desulfurization facility at an oil refinery plant. The other is an example of life cycle planning applied to valve actuators of air-conditioning units, in which reuse is integrated in maintenance. 4.1. Operation and maintenance integration for the direct desulfurization facility at an oil refinery plant [9] Deterioration modes of the direct desulfurization facility In the direct desulfurization facility, heavy oil is heated by the furnace and desulfurized through a decomposition reaction at high pressure and high temperature in the presence of a catalyst. To keep the oil temperature at the required value, the external temperatures of the heating tubes should be increased depending on the amount of carbon deposit on the inner walls of the tubes. The increase in the external temperatures of the heating tubes, however, accelerates the progress of creep damage and reduces the life of the tubes. Therefore, it is necessary to execute decoking by shutting down the facility. Since shutdown of the facility induces huge production losses, we need to consider its effect on operations when determining the proper maintenance plan. To evaluate expected losses, we construct a mathematical model of the various deterioration modes occurring in the facility, such as catalyst degradation, carbon deposit on heating tube inner wall, heating tube creep damage, wastage owing to metal dust corrosion of heating tubes, brick separation of the wall of the combustion chamber, and burner tile breakage and nozzle clogging. Conducting life cycle simulation Using these deterioration models, we conducted LCS to evaluate operations and maintenance losses. The effects items are identified for three objects: humans, facilities, and the environment. All items are monetarily evaluated in this case study. For facilities, we identified the losses caused by operation and maintenance actions. Regarding maintenance, we consider the labor cost and cost for parts necessary for maintenance actions such as inspection, repair, and replacement, as well as the production losses induced by maintenance actions. Regarding operations, production losses owing to reduced production volume and excess fuel costs are considered. In addition, life shortening of the facilities is considered as a depreciation expense loss, which is related to both maintenance and operations. For humans and the environment, we consider compensation. We set the maximum acceptable risk to manage failure that could lead to catastrophic damage. For the direct desulfurization plant, creep rupture of the heating tubes could cause fatal damage. Therefore, the allowable rupture probability is set at 0.0001(month-1), which corresponds to 81.3% of the life consumption rate of the creep life. Preventive maintenance, that is, tube replacement, is executed

29


when the life consumption rate reaches this limit in the simulation. With regard to the maintenance policy for each deterioration mode, we apply those adopted in the current practice. All maintenance actions are executed during shutdown maintenance (SDM) except decoking. Since creep rupture causes fatal damage, we assume that the necessity of decoking is determined in each term and that it is executed even between SDM periods. The SDM interval is changed from 10 to 21 terms. The maximum maintenance cycle of 21 months is determined because catalyst degradation has a huge impact on the life expectancies of the heating tubes when the cycle becomes longer than 21 months. The TBM interval for each deterioration mode is changed from 8 to 35. Regarding operations, we set the allowable range of production volume from 4800 kl to 6400 kl per day. Since the facility’s characteristics mean that carbon deposition is accelerated in the case of low production volumes, we set the lowest production volume limit. Simulation results The results of the simulation are shown in Figures 10 (a) and (b), in which the average losses per unit term, the average marginal profit per unit term, the lifetime of heating tubes, and the holding risk per unit term are indicated. Figure 10 (a) shows that most losses in this setting are induced by production losses and the depletion of heating tubes. It shows that if we prioritize maintenance and shorten the SDM cycle, production cost increases. On the other hand, if we prioritize operations and lengthen the SDM cycle, emergency maintenance losses and depreciation cost increase. The total losses have the minimum value at the SDM cycle of 15 months, while the marginal profit has the maximum value for this cycle. However, Figure 10 (b) shows that the holding risk increases monotonously as the SDM cycle increases, whereas the lifetime of tubes decreases. The amount of risk shown in Figure 10 (b) corresponds to emergency maintenance losses.

Valve actuators are widely used in the air-conditioning systems of buildings in which each floor is air conditioned independently by air-handling units on each floor. An airhandling unit exchanges the thermal energy of water with that of air. Valve actuators are used to control the water flow. It is important to maintain their functions in order to ensure that the air-conditioning system operates appropriately. A valve actuator consists of a valve part, which adjusts the water flow, and an actuator part, which drives the valve according to the control signals provided by a controller. TBM is the maintenance policy usually applied to valve actuators. In principle, valve actuators are replaced after 15 years of use. High reliability can be achieved with this maintenance policy. However, such a maintenance policy has been adopted on an empirical basis, and its effectiveness has not been evaluated quantitatively. This policy may result in excessive maintenance and heavy LCC and environmental load. Therefore, we attempt to improve maintenance efficiency by applying the life cycle maintenance–planning method. Evaluation of maintenance plans In order to develop a better maintenance plan, we evaluated six maintenance plans, including the currently used plan. The first four plans adopted TBM with different maintenance cycles of 10, 15, 20, and 25 years. The other two plans adopted CBM and BM. In the case of CBM, the valve actuators are inspected every year. When a symptom of a Creep rupture Brick separation Burner nozzle clogging Burner tile breakage

13.4

Marginal profit

3.5

13.2

3.0

13.0

2.5

12.8

2.0

12.6

1.5

12.4

1.0

12.2 10 11 12 13 14 15 16 17 18 19 20 21

Valve actuators of air-conditioning systems

1.8

900 800

Lifetime of heating tube

700

Average losses and average marginal profit

1.4

600

1.2

500

1.0

400

0.8

300

0.6

200

0.4

100

0.2 0.0

0 10 11 12 13 14 15 16 17 18 19 20 21

SDM cycle [months]

(a)

1.6

SDM cycle [months]

(b)

Life time of heating tube and holding risk

Fig. 10 Simulation results of operations and maintenance integration for direct desulfurization facilities of oil refinery plants

Total holding risk [100 million yen/month]

Losses [100 million yen/month]

4.0

4.2. Life cycle planning for the valve actuators of airconditioning units [10]

Marginal profit [100 million yen/month] Lifetime of heating tube [months]

Production losses Depletion of heating tube Fuel cost losses Preventive maintenance costs Emergency maintenance losses

Therefore, applying better maintenance policies to reduce the emergency maintenance cost contributes to an increase in the marginal profit. The results show that selection of the proper SDM cycle can increase the total profit. At the same time, we need to pay attention to the increase in fatal risks at the facility.

Cost [yen / year]

failure rate

0.025

4000

0.020

3000

0.015

2000

0.010

1000

0.005

0

10

15

20

TBM

0.000

25

5000

Cost [yen / year]

Cost

5000

Cost

failure rate

0.025

4000

0.020

3000

0.015

2000

0.010

1000

0.005

0

CBM BM

TBM TBM+ Reuse (15-year (15-year cycle) cycle)

BM

Failure rate [ / year]


Failure rate [ / year]

30

0.000

Fig. 11 Cost and failure rate of different maintenance plans

Fig. 12 Simulation results showing the effects of reuse

failure is found in a valve actuator, the valve actuator is replaced. By contrast, in the case of BM, the valve actuator is replaced only upon its failure. The results of the evaluation are shown in Figure 11 in terms of the cost per year, which is the sum of the effects on the running cost of building owners and maintenance cost of valve actuators. Figure 11 also shows the mean failure rate. This figure shows that in the case of TBM, the longer the maintenance interval is, the lower is the LCC and the higher is the failure rate. However, in the case of BM, the cost is lower and the failure rate higher as compared to TBM. In the case of CBM, both the cost and the failure rate are high, because it is difficult to detect the symptoms of failure without disassembling the valve actuators in the case of the current design. From the viewpoint of LCC, BM appears to be the best maintenance policy according to the evaluation results. However, we may need to consider the effect of low reliability on brand loyalty. Therefore, we estimate the increase in the accumulated sales loss arising because customers switch brands owing to frequent valve actuator failures. We assume that a customer switches to another brand when more than two failures happen in one building per year. (Buildings in Japan have an average of 40 valve actuators.) The result shows that the sales loss is considerably larger than the cost reduction with BM, if we consider the LCC of more than 20 years. None of the five other maintenance plans provides a better solution in terms of cost and reliability than the currently used plan.

cleaning and replacing the sheet ring and the gland packing. The rest of the valve consists of cast iron and stainless steel, which are quite durable. Therefore, we assume that a valve can be remanufactured and reused two times. This means that the main body of the valve can be used for 45 years without decreasing the reliability of the valve. Reuse of the actuator and the valve is effective for reducing cost. We estimate that the remanufactured actuator and valve cost 40% and 50% less than the newly produced ones, respectively. The effect of reuse is evaluated in terms of cost and reliability. The result is shown in Figure 12 along with the results obtained in the cases of only TBM and BM. The figure indicates that the proposed maintenance plan when incorporated with reuse brings about a cost reduction of approximately 40% as compared to the present maintenance plan, while maintaining the same level of reliability. In addition, the environmental load can be also reduced by 50%.

Integrating reuse option into the life cycle plan In order to improve the life cycle of valve actuators in terms of cost and reliability, we propose the adoption of the reuse option, which is incorporated into the maintenance plan. Currently, valve actuators are periodically replaced every 15 years. However, the lives of valve actuators are determined by the component with the shortest life, even though other components have longer lives. In the case of actuators, the potentiometer has the shortest life and is therefore the critical component. Other components in the actuator have considerably longer lives. Therefore, we assume that the actuator can be remanufactured by replacing the potentiometer and can be reused for another 15 years without decreasing its reliability. In the case of a valve, leakage owing to the deterioration of the sheet ring and the gland packing and seizure of the ball owing to the presence of foreign objects are major deterioration and failure modes. These can be reset by

5. Conclusion To realize a sustainable society, it is indispensable to maintain and improve the functions obtained by using various artifacts with minimum environmental load and resource consumption. For this purpose, we need to realize true circular manufacturing instead of just adding reverse flows to the conventional manufacturing systems. The core technologies of circular manufacturing are diagnosis, restoration, and upgrading technologies that have been studied in the context of maintenance engineering. Another important technology for circular manufacturing is that used to deliver artifacts to the right users at the right time depending on the physical and functional state of the artifacts and the users’ requirements. Such technologies have been discussed in the context of reuse and remanufacturing. Therefore, we proposed the concept of MCCM, in which maintenance technology is viewed as a core technology, and reuse and remanufacturing technologies are integrated with it. In this paper, we discussed maintenance and reuse management technologies in the context of circular manufacturing. Although there is much to do to realize such a concept, such as transforming business models so that the delivery mode of artifacts can be adapted to circular manufacturing and the innovation of technology that enables the personalized restoration of artifacts, the author hopes that this article would be of some help in moving forward with a manufacturing paradigm change.


References [1] Takata, S., Kimura, F., van Houten, F. J. A. M., Westkamper, E., Shpitalni, M., Ceglarek, D., Lee, J., Maintenance: Changing role in life cycle management, CIRP Annals—Manufacturing Technology, Vol.53/2, 2004, p.643-655. [2] Umeda, Y., Takata, S., Kimura, F., Tomiyama, T., Sutherland, J. W., Kara, S., Herrmann, C., Duflou, J. R., Toward integrated product and process life cycle planning—An environmental perspective, CIRP Annals—Manufacturing Technology, Vol.61/2, 2012, p.681-702. [3] Ni, J., Jin, X., Decision support systems for effective maintenance operations, CIRP Annals—Manufacturing Technology, Vol.61/1, 2012, p.411-414. [4] Chang, Q., Ni, J., Bandyopadhyay, P., Biller, S. Xiao, G., Maintenance opportunity planning system, ASME Journal of Manufacturing Science and Engineering, Vol.129, 2007, p.661-668. [5] Umeda, Y., Kondoh, S., Sugino, T., Yoshikawa, H., Analysis of reusability using “marginal reuse rate,” CIRP Annals—Manufacturing Technology, Vol.55/1, 2006, p.41-44.

31

[6] Takata, S., Tsubouchi, K., Yoshijima, S., Exhausting lifetime by multitiered reuse with forced circulation in closed-loop manufacturing, International Journal of Sustainable Manufacturing, Vol.1/4, 2009, p.450-462. [7] Takata, S., Tsubouchi, K., Maximizing utilization rate of office automation equipment by intraoffice circulation, CIRP Annals— Manufacturing Technology, Vol.58/1, 2009, p.33-36. [8] Urano, K., Takata, S., Module reconfiguration management for circular factories without discriminating between virgin and reused products: Reengineering manufacturing for sustainability, Proc. of the 20th CIRP Conference on Life Cycle Engineering, 2013, p.603-608. [9] Tsutsui, M., Takata, S., Life cycle maintenance planning method in consideration of operation and maintenance integration, Production Planning and Control, Vol.23/2-3, 2012, p.183-193. [10] Momiyama, T., Takata, S., Life cycle maintenance in combination with reuse—Application to valve actuators for reducing life cycle cost and environmental load, Proc. of the 17th CIRP Conference on Life Cycle Engineering, 2010, p.306-311.