What type of decision is the following: did the amount of daily transactions increase?

3.2 Tactical and operational decisions

Tactical and operational decisions are the medium- and short-term decisions which can be alternated annually, weekly, or even daily (Awudu and Zhang, 2012). They are made under the constrained structure of strategic decisions that are typically made before investigating tactical and operational decisions (De Meyer et al., 2014). Compared with strategic decisions, tactical and operational decisions are typically made at a smaller scale(e.g., biorefinery level or process level). Tactical and operational decisions may include the following aspects:

Production planning determines detailed design and operations of unit processes included in BSC, such as supplying biomass and other raw materials (Zhang and Hu, 2013), process design (Tong et al., 2013; Kazemzadeh and Hu, 2013), and scheduling (Sharma et al., 2013; Beamon, 1998).

Inventory planning determines the quantity and timing of materials or goods in stock (Cachon and Fisher, 2000; Stadtler, 2005; Min and Zhou, 2002) which needs to be aligned with production capacity, fuel distribution, and biomass supply (Tong et al., 2013; You and Wang, 2011; Azadeh et al., 2014). The storage contains raw materials for manufacturing, intermediate productions, and final product for distribution.

Logistic management refers to managing and implementing the sufficient and effective flows of materials (e.g., raw materials, intermediates, and products), goods, and information from original suppliers to end users (Bowersox, 1997; Lambert and Cooper, 2000; Sokhansanj et al., 2006).

Fleet management decides the movements of materials between different BSC stages (Awudu and Zhang, 2012). Fleet management plays a crucial role in BSC as it directly affects the robustness of the transportation network (Ravula et al., 2008; Eriksson and Björheden, 1989; Van Wassenhove and Pedraza Martinez, 2012; Thomas and Griffin, 1996).

Depending on the predetermined strategic decisions and the objectives of BSC, different tactical and operational decisions mentioned were considered in previous BSC cases (see Table 10.2). Zhang and Hu (2013) built two models to optimize the strategic decisions of facility locations and capacity, then tactical and operational decisions such as monthly biorefinery production planning and inventory control were investigated and determined. Some studies developed optimization approaches to make strategic and tactical decisions simultaneously. For example, Lin et al. (2014) established a model to optimize the large-scale biomass-to-ethanol SC where the strategic (e.g., farm and facility locations and capacities) and tactical decisions (e.g., biomass production planning, plant operating schedules, and inventory control) were optimized simultaneously. An et al. (2011b) established a model considering multiple types of lignocellulosic biomass and the material flows in the BSC. This model could be used for both strategic and tactical decisions including facility locations and capacities, technology types, production plans, transportation strategies, and storage amount. Ekşioǧlu et al. (2009) integrated the long-term decisions (e.g., capacity, location, and the number of biorefineries) and mid-term logistic decisions (e.g., biomass supply) (Ekşioǧlu et al., 2009). As a BSC usually has a large number of components, determining tactical and operational decisions without considering the uncertainty related to each component may lead to poor performance of the whole SC (Awudu and Zhang, 2012). Different approaches (e.g., stochastic programming and fuzzy logic) have been developed to model and address uncertainties in BSC design, which are further discussed in the following sections.

Abstract

Tactical and Operational decision making requires considering a large amount of data, which must be properly stored and interpreted. In this context, an on-line information system has been developed to allow the integration of real time reactive tools, which can constitute a useful operator support in the event of process and/or scenario disturbances. The use of this integrated system as a training tool over a simulated Supply Chain scenario has proved to help the Process Systems Engineering students to improve the information comprehension capabilities, as well as to enhance the way they apply their knowledge to plant and process optimization.

8.4 Conclusions

Tactical decisions in the supply chain, which are medium-term decisions that typically span from a few months to one year, identify medium-term policies associated with different processes of the supply chain, according to the network structure established in the strategic-level decision-making. This chapter provides a framework for the tactical decisions that are made at different stages of biomass-to-biofuel supply chains, and discusses uncertainties in biomass supply chains, including biomass yield, biofuel demand and price, and biomass conversion rates. Biofuel supply chain master production planning seeks to provide a production plan to produce biofuels by utilizing the available conversion capacity of biorefineries in a timely and cost-effective manner. To illustrate how to provide an optimal planning model for biofuel supply chains, a multiperiod MILP model is presented in this chapter to address the master planning of the JCL-to-biodiesel supply chain, which determines the optimal values of tactical decisions in each time period. The deterministic biodiesel supply chain master planning model is then extended to a robust counterpart model, taking into account uncertainty in JCL yield, JCL production and preprocessing costs, oil extraction costs, biodiesel production cost, inventory holding cost of materials at different facilities, transportation cost, as well as biodiesel demand and price. The robust possibilistic programming is adopted to deal with uncertain parameters that are modeled as fuzzy numbers.

To address the problem of biorefinery process synthesis and design, a biorefinery superstructure model for biodiesel production from microalgae is proposed, which determines the optimal production pathways while taking into account uncertainties related to technical factors. To cope with uncertainty in the yield parameters of the proposed model, robust scenario-based stochastic programming is applied. The results of the sensitivity analysis based on a deterministic version of the model show that varying the percent yield of reactions occurring in the processing units has a significant impact on both the design and profitability of the microalgae-based biorefinery. However, the accurate estimation of the percent yield, which is defined as the ratio of the actual yield to the theoretical yield of a reaction, is difficult, particularly in the case of biofuel production from microalgae because there is at present no commercial-scale microalgae biorefinery.

1 Introduction

The tactical decision making problem has been successfully studied in the last 20 years; recent approaches intend to integrate typical planning models for novel Process System Engineering (PSE) applications, covering a significant number of issues to be considered in practice, like uncertainty (Balasubramanian et al. 2002), Multi-objective optimization (Bojarski et al. 2009), integration of different decision making levels (Sung and Maravelias 2007), etc. but most of them disregard the effects of interaction (cooperation and competition) among SC's.

But the incorporation of information from the different interacting enterprises in a single process model is essential to rationalize the tactical management of any of these interacting systems, requiring the procurement of RM from different suppliers, the allocation of materials to different plants, or deciding the distribution of products to the final consumers. So the features of each one of the single echelons of the entire SC have to be integrated in a specific model with its own objectives and management practices.

One clear example of this need can be found associated to the field of energy management, especially in “green” energy generation, by integrating in a competitive scenario biofuels gasification and combustion processes, and the exploitation of other energetic resources. The design, planning, and operation decision making of energy networks arise as new challenges for the PSE community (Perez-Fortes et al. 2011 Zamarripa et al. 2011).

This work proposes a mixed integer non-linear programming (MINLP) model able to optimize the overall SC planning of SC in order to deal with the complexity arising the consideration of multiple supply chains with independent objectives, the coordination of requirements and their integration with green energy generation systems,.

The proposed solution is based on modeling the main characteristics of multiple echelons (including suppliers, production plants, storage centers, waste, and markets) considering the behavior of each echelon. The resulting model is flexible enough to optimize the overall total cost of the entire SC (Suppliers SC, production SC, waste SC, etc.). Furthermore, it is capable to examine and compare different integration options of different sub-networks coordinating the inputs and outputs of each part.

3.3.2 Biomass storage planning

One of the main tactical decisions associated with biomass storage is determining the seasonal or monthly inventory level of biomass in each storage center. The motivation for storing biomass is to balance its supply and demand because there is high regional and seasonal variability in biomass production while the demand for biofuels and biomass feedstocks must be satisfied continuously all year round. Storage decisions are addressed by multiple-period supply chain planning models where the inventory levels in consecutive periods are optimized such that the continuous supply of biomass is ensured.

The selection of biomass storage method is another decision that needs to be made in the medium-term planning level. Various methods have been proposed in the literature for storage of biomass, among which ambient storage covered with a plastic film is the cheapest option. However, the drawbacks include the lack of control over biomass moisture, significant material loss, and possible risks for human health. To avoid biomass quality degradation because of fermentation, infections, and material loss, more expensive methods such as closed warehouses with hot air injection and covered storage facilities with a pole-frame structure can be used (Rentizelas, Tolis, & Tatsiopoulos, 2009). It is clear that the tradeoff between material loss rate and storage costs must be considered in selecting the appropriate storage method.

3.2.1 Optimization of early-phase testing

This subproblem encompasses the tactical decisions concerned with the planning of all activities that are required to be carried out in a clinical trial from active ingredient manufacturing, through drug product manufacturing, packaging, and distribution to the clinical sites, and finally to drug dispensation to patients at clinical sites.

This kind of problem has received attention from the PSE community utilizing previous works from the process planning and scheduling area. Schmidt and Grossmann (1996) proposed various MILP (mixed-integer linear programming) optimization models for the scheduling of testing tasks with no resource constraints with a discretization scheme in order to induce linearity in the cost of testing. Jain and Grossmann (1999) extended these models to account for resource constraints. Subramanian et al. (2003) proposed a simulation-optimization framework that takes into account uncertainty in duration, cost, and resource requirements and extended this model to account for risk. Maravelias and Grossmann (2001) developed an MILP model that integrates the scheduling of tests with the design and production planning decisions with a two-stage stochastic optimization approach that is adapted to account for the uncertainty in the outcome of the trials. The model supports decisions related to the choice of products be tested, assignment of test resources, and task outsourcing. A heuristic algorithm based on Lagrangian decomposition was developed for this purpose.

Military training

Here the dramatic use of video to portray the results of tactical or operational decisions is obvious. Bad military decisions costs lives, so realistic simulation is even more important as a training medium than in most of the cases shown above. In fact, military training will encompass operator training for weapons, tanks and engineering applications of equipment, together with battlefield simulations for operators. It will use all the maintenance applications listed so far, take up the management training requirements for interpersonal skills, leadership, motivation and interviewing and apply them both to peacetime administrative situations and to wartime field scenarios. Military training, in a cost-effective and non-hazardous fashion, will be one of the most powerful applications of interactive video.

2.3.4 Agent Definition and Behavior

The key agents in the model are the power-producing companies. Their tactical decisions consist of offering their output to the power market. They bid their power based on marginal costs, which includes the cost of CO2 emissions in the cases with a carbon policy. Thus markets are simulated in which power producers negotiate the electricity, fuel, and CO2 prices. Their strategic decisions cover investment in and decommissioning of power plants. Each agent's decision process is as follows:

Decide per power plant whether it should be dismantled. The decision to dismantle is taken when the technical lifetime of a power plant has expired (after 20 years for wind farms, 30 years for gas and coal plants, and 40 years for nuclear) or if the plant caused continuous operational loss for over 5–9 years.

Estimate whether there is a need for new generation capacity in 3 years. The estimate of the demand for capacity in 3 years is based on an extrapolation of the electricity demand trend of the past 3 years. Capacity expansion decisions take into account investments and decommissioning already announced by competitors. Continuous operational losses will cause unannounced decommissioning; thus agents' planning is not perfect and investment cycles can occur. Limited overinvestment is modeled to dampen those investment cycles.

If Step 2 results in an investment decision, the agent needs to select a technology for its new plant. Its decision is based on the life-cycle cost per MWhe produced. The life-cycle CO2 cost is based on current CO2 taxation levels or, under emissions trading, the three-year average CO2 auction price. The total life-cycle cost must be recovered by electricity income or else the investment is cancelled. In the latter case, another agent will get the opportunity to invest. The order in which the agents make their investment decisions varies randomly. In addition to financial aspects, an agent's conservativeness, aversion to nuclear power, and risk attitude affect its decisions. Despite the large weight of financial considerations, these individual style aspects have an effect, especially when financial differences between options are small. Conservativeness is modeled as “preferring more of the same”; risk attitude translates to different responses to historic variance of CO2 and electricity prices.

In the case of emissions trading, each year electricity-producing agents complete the following actions concerning the operation of their power plants:

Purchase emission rights in the annual auction. The auction bids are based on the “willingness to pay” per installation, which is determined as the expected electricity price less the marginal costs of each unit, divided by the CO2 intensity. The bid volume equals the expected electricity sales volume times the CO2 intensity of the power plants that are expected to be in merit.

Offer electricity to the market (which is modeled as a power pool). Each plant's capacity is offered at variable generation cost (fuel cost, variable operating and maintenance cost, and CO2 cost). The CO2 costs of a generator equal the CO2 price times its CO2 intensity. In case insufficient CO2 rights have been obtained, CO2 cost is equal to the penalty for noncompliance.12

Acquire the required amounts of fuel from the world market, which are calculated from the actual production and fuel usage.

Bank surplus CO2 rights or pay the penalty in case there is a shortage of CO2 rights. Surpluses and shortages are calculated from the actual production levels and the volume of emission rights owned by the agent.

A difficulty with this procedure is that the CO2 and electricity markets are mutually dependent. Though there is only one CO2 price per year, the use of a load-duration function with 10 steps means that 10 different electricity prices are developed for each year. Therefore we need to model arbitrage between these 10 periods. Because the demand for CO2 credits is different at every step in the load-duration curve, we had to develop an iterative process in which arbitrage between the demand for CO2 in these markets takes place in such a way that total annual demand for CO2 satisfies the emissions cap and a single annual CO2 price develops. We adopted the following procedure: Since the outcome of the CO2 market is input to the power market and vice versa, Steps 1 and 2 are computed via an iteration that is complete when stable prices have been established for the entire year. In each simulation interval, we start with the prices of the previous year. In each iteration, first the CO2 auction is cleared (Step 1), which results in a CO2 price. This price is then used to calculate power market offers, and for each of the 10 sections of the load-duration curve this market is also cleared (Step 2). This new clearing price for electricity is fed into the bids for the CO2 auction as the expected price of electricity (Step 1), and so on. Upon completion of this iteration, emissions trading (Step 1) has effectively been completed.

Under carbon taxation, Step 1 is skipped; in Step 2, the CO2 cost is the carbon tax times the CO2 intensity. In this case, the CO2 price is exogenously determined. The electricity market bids simply incorporate this price. No iteration is necessary. Step 4 is replaced by paying carbon tax to the government. In case there is no carbon policy, the calculation procedure consists of Steps 2 and 3, with a CO2 price equal to zero.

Where Tactical Decisions Are Made

When considering business value, it is worthwhile noting that there is a symbiotic relationship between tactical decisions and strategic decisions. Fig. 16.1.8 depicts this relationship.

Fig. 16.1.8 shows that—for the most part—tactical activities and transactions occur in the application environment. And strategic activities and decisions occur in the data warehouse and the data mart environment. A cycle of processing occurs between the two.

Transactions are run in the application environment. Data are produced as a by-product of those transactions. The data find its way into the data warehouse. The data in the data warehouse and the data mart are studied, and new decisions are made. The new decisions have an impact on the transactions that are run. In turn, a new set of transactions are run, and their data are stored in the data warehouse.

This cycle of data has a profound impact on the business of the corporation.

Abstract

Decisions in biomass-based production chains can be classified into three main categories; strategic, tactical, and operational decisions that cover respectively long-, mid- and short-term design, and management and planning choices about supply chain configuration and operations. To deal with large-scale and complex problems effectively and to achieve an integrated, robust and reliable supply chain, numerous decisions at different levels should be handled simultaneously in design and planning practices. These decisions may be monitored, revised, repeated, or changed in different time spans, which specifies the type, scope, and level of the decisions. This chapter focuses on the decision levels in design, planning, and management of biomass-based production chains as well as the main economic, environmental, social, and technical objectives and criteria captured at all decision levels. In addition, the multiple attribute and multiple criteria decision-making methods to capture the explained objectives and criteria in tandem are described.