Summarize

1. Conclusion, suggestions and managerial implications

The focus of this study is on CLSCC for new product and its reconditioned version. Our integrated CLSCC optimization model simultaneously determines optimal production/sales plan as well as configuration of entire CLSC. We conduct a real-world case study of a battery manufacturer based in India to test our model. The results and analysis of our optimization model are summarized below. First, our results indicate that our integrated CLSCC model performs Better than firm's existing supply chain design. Our results indicate That it is possible to generate higher sales revenue by choosing alter- Native suppliers with shorter lead time than the current suppliers. However, the firm decides to continue its long-term relationships with The current suppliers due to the existing trust. Also, the management Of Xtra Power believes that building reliable relationships with the new Suppliers is risky proposition. The management is apprehensive that any change will increase its costs and hence decrease the profit. It is advisable to assess the potential costs and profits relevant to the new suppliers. The firm should take into consideration other benefits, like reduction in lead time, reduction in inventory and etc., which may be more than the higher costs to be incurred. In addition, we solve our optimization model using data gathered from the literature which is aggregate of the data collected from various firms in the US and other countries(see Table A-1in Appendix). These results indicate that CLSCC model incurs an additional cost of 2%, but net profit is increased by 25%. Whereas, the data obtained from Xtra Power results in an additional cost of 86% and 17% increment in net profit. We agree that benefit-cost ratio reduces in the case of Xtra Power, but we cannot generalize it as it is case specific.

Second, Xtra Power decided to pay13–15%of new battery's price toacquire used battery, to match the rate of competitors. Our modelprovides a threshold acquisition rate for a particular combination ofsales price ratios. The threshold rate can be used to determine themaximum acquisition price that a firm could pay to acquire usedbatteries ,at which CLSCC stays profitable. Beyond this threshold rate,FSCC model becomes more profitable than CLSCC model, indicating That entering reconditioning business is not profitable for the firm.Our results indicate that threshold rate is almost twice or more thantwice as compared to firm's chosen rate (13–15%) for all analyzedscenarios. After reviewing our analysis, Xtra Power decided to increase the acquisition rate in order to receive sufficient number of used batteries to justify the higher cost of a profitable CLSCC.As shown in , [71] setting an appropriate price corresponding to a specific quality grade would not only generate profits but also reduce inventory. While fixing the acquisition price the firm should also consider Combination of sales-price ratios of new and reconditioned batteries. Third, literature suggests that higher return rate would fetch more profits than lower or moderate return rates. However, our results indicate that it is not always true. The combination of sales price ratios of new and reconditioned batteries decides the relation between return rate and profit. While fixing the sales-price ratios of new and reconditioned batteries, the firm should consider the potential return rates of the used batteries since it has significant effect on total profit.

multiple options selection at each stage would be an interesting extension. Second, we assume that firm receives a proportion ω (0rωr1)of new product's sales, in a single period, after their useful Life τ. It may also happen that a firm may receive product returns any Time after new product's sales. There fore,to model and solve such case Deserves further attention. Third, we consider distinct markets for new and reconditioned products. This assumption may not be valid for commercial products whose buyers purchase the product based on its functionality[56,73].Future studies can relax this assumption. Fourth, we assume a sizeable demand for reconditioned products. It May also happen that there is a limited demand for reconditioned products. There fore,it is desirable to determine the approximate demand for reconditioned product. Fifth, we assume all returned used products would be reconditioned and sold in the same period. In response to demand fluctuation, a firm may decide to carry and sell the reconditioned products over multiple periods. There fore, this assump-tion can be relaxed in the future studies. Sixth, we assume same quality for all accepted used products. Model different quality grades for acceptance would be an interesting extension of the current study. Seventh, our study ignore potentials competition. Although it is difficult

To model competition in supply chain configuration problem, yet there Lies an opportunity to do competitor analysis. Finally ,in our model we consider single product, whereas[11]configure supply chain for plat- form products but only for FSC. Future studies may extend our mod-eling approach to consider plat form products for CLSCC.

8. Limitations and scope for future research

The limitations of this study are as follows. First, we assume single Sourcing for each stage of CLSCC. While single sourcing has its own Merits including better information sharing, lower cost, better quality, and focused investment[15,72],but it is prone to certain risks and disruptions. Extension of our CLSCC modeling approach to consider

2.Acknowledgment

This paper is developed based on the first author's doctoral dissertation. We would like to thank Dr. Lev, Editor-in-Chief, and also the two reviewers for their thoughtful comments. Their comments have significantly helped us to improve the content and presentation of this paper. Also, we thank the executives at the Xtra Power Auto Industries for their generous assistance in devel-opment, data collection and results validating of the case study

presented in this paper. In addition, we would like to acknowledge Fogelman College of Business and Economics, enterprise Simulation and Optimization Lab (eSOL) and FedEx Center for Supply Chain Management, for partially supporting the second author's efforts in

completing this research study.