Parameter Prioritization in Techno-economic Model for Decision-making in the Petrochemicals Industry

2.1. Background of the Malaysian Economy and the Petrochemicals Industry

Malaysia's ethylene market is dominated by High-Density Polyethylene (HDPE), Low-Density Polyethylene (LDPE), and Ethylene Oxide (EO). Since Vinyl Chloride Malaysia (a PETRONAS subsidiary) exited the vinyl business in late 2012, there has been no more ethylene demand for EDC and VCM. The recent drop in ethylene demand is also due to HDPE losing its market competitiveness. Although local demand for HDPE continues to grow, limited production capacity has resulted in increased imports from the Middle East to meet the demand gap. Currently, downstream production will increase as a result of a more competitive ethylene supply from the expanded cracker, followed by a 70% increase from the Pengerang Integrated Complex (PIC). The complex will have the potential to generate one million tonnes of ethylene per year and will also have HDPE and Linear Low-Density Polyethylene (LLDPE) production capabilities [10]. In Malaysia, PETRONAS Chemical Ethylene Malaysia, Lotte Chemical Titan, and PETRONAS Chemical Olefins Malaysia operate ethylene crackers. Lotte Chemical Titan, located in Pasir Gudang, is the largest manufacturer, with a combined capacity of 735,000 metric tonnes per year. Lotte Chemical Titan expanded its ethylene capacity by 90,000 tonnes per year in August 2017, after Malaysia's ethylene capacity had been stable for a long time. This expansion utilized catalytic cracking, which was combined with the existing naphtha thermal cracker to create the world's first hybrid ethylene-producing facility. The massive new PETRONAS refinery and naphtha-based steam cracker project with a design capacity of 1.3 million metric tonnes per annum of ethylene is more substantial than the polyethylene, polypropylene, and Mono Ethylene Glycol projects (MEG) [10].

Polypropylene accounted for over 55% of Malaysia's propylene demand, with oxo-alcohols and acrylic acid also being significant consumers. Since 2008, Malaysia's annual propylene capacity has been slightly over one million metric tonnes. Nonetheless, the market for propylene in Malaysia will experience substantial shifts during the forecast period. In August of 2017, Lotte Chemical Titan expanded its naphtha cracker, increasing annual propylene capacity by 160,000 metric tonnes, while the steam cracker in Pengerang will produce roughly 660,000 metric tonnes of propylene and polypropylene per year [11]. The first butadiene rubber plant in Malaysia was inaugurated at the end of 2015, with an annual capacity of 50,000 tonnes of butadiene, or roughly a quarter of historical usage levels [12]. In recent years, the rise of NBR latex production has been the key demand driver for butadiene in Malaysia. Malaysia is the world's leading provider of natural rubber and NBR gloves, with a 60% market share. In recent years, the demand for NBR gloves has steadily outpaced that for natural rubber gloves [13]. Butadiene consumption in NBR latex will continue to fuel expansion, as rising health care expenditures, ageing populations, and higher global cleanliness standards will increase the need for medical gloves [14]. By the third quarter of 2018, Synthomer increased its capacity by 90,000 tonnes per year, bringing the total to 270,000 tonnes per year [15]. This expansion will push local demand for butadiene up to 30,000 tonnes. Another contributor is Toray Plastics; their recent expansion means ABS is now a small but growing part of butadiene usage here. Earlier in March 2008, Titan Petrochemicals started its 100,000 tonne per year butadiene plant, before the whole company was bought by Honam Corporation from South Korea. This plant uses “mixed C4” from its own steam cracker, which used to be exported. Currently, about 50% of the butadiene produced is sent to Kumho Petrochemical in South Korea. In 2020, PETRONAS began supplying 1.3 million tonnes of feed yearly to the RAPID naphtha-based steam cracker. This complex includes a distillation plant that can produce 190,000 tonnes of butadiene every year [16].

Many factors have affected the performance, profit, and investment levels of the global petrochemical industry. This includes environmental rules, new technology, changes in what customers want, and global politics or economic issues.

For Malaysia, one of the biggest challenges is the shortage of resources. It is very important that our offshore fields can still supply oil and gas feedstock at a lower price compared to other countries. This low cost is what allows Malaysia to produce petrochemical products at a competitive price. To stay ahead, Malaysia must make sure these prices remain low. In terms of total capacity, Malaysia’s crude oil reserves rank 28th in the world (573,000 barrels per day), while our natural gas reserves rank 12th (74.2 billion cubic meters) [17]. Malaysia also boasts the world's largest LNG production facility, with an annual output of 29.3 million metric tons, located on its territory [16]. Despite having such a large oil reserve, Malaysia is actually a net importer of the commodity due to falling domestic production in recent decades. Malaysia also has some of the world's most advanced oil refineries, with a combined daily output of 955 thousand barrels from its five operational refineries [17].

The petrochemical industries must use robust technology to optimize the available resources. Companies should also keep looking for new technology and inventive applications for the ones they currently have. If the petrochemical sector does not evolve in this manner, it will fall behind the market and lose any advantages it may have. The necessity for qualified labor capable of handling the complexities of the petrochemical business should also be considered. Nevertheless, basic training may not be enough to provide the company with a competitive advantage over competitors who may have more qualified employees. Lastly, the environment is severely impacted as a result of carbon dioxide overproduction, and the petrochemical industry is responsible for the majority of these emissions. It will be critical for industry players to solve it by redesigning new and existing operations to produce less carbon dioxide [18].

2.2. Background of Selected Petrochemicals Production

Propylene oxide is a component in the synthesis of polyether polyols, propylene glycols, and other specialized chemicals. In the 1970s, Chlorohydrin (CHPO) or chlorohydrin with chlor-alkali (CHPO CA) was at the height of its popularity. Essentially, this technique has not been widely adopted in most industrialized nations because of the environmental concerns around chlorine byproducts. The chlorohydrin process still accounts for over 40% of global capacity as of 2016, despite its declining popularity and multiple closures [19]. Other technologies that have been developed by the researchers as substitutes for the chlorohydrin process include Propylene Oxide-Styrene Monomer (POSM) and Propylene Oxide-Tertiary Butyl Alcohol (POTBA) processes. The POSM process started in 1974 because it was a cheaper way to make propylene oxide. Today, it is the second-most used method in the world. Later in 2003, Sumitomo Chemical came out with a new process using cumene hydroperoxide (CuPO). This method is advantageous because it does not produce any extra co-products.

Other companies like Evonik, Dow, and BASF also found a way to make propylene oxide using hydrogen peroxide (the HPPO process). These companies started looking at each other’s work in 2002 and began working together the year after. They eventually built a full-scale HPPO plant at the BASF site in Antwerp, Belgium [19].

2.3. Objective Weighing Method

Objective weighing method is a versatile and effective tool for decision-making, data analysis, and information theory [9, 20]. One of the primary advantages of this method is its ability to provide an objective way to assign weights to different criteria or features based on their variability or uncertainty. This will aid in minimizing the bias during the weighting process, and the weights will not be based on individual opinions or random decisions. The other important advantage of the objective weighing technique is that it deals with uncertainty. This method can be used to provide a more suitable distribution in situations of high uncertainty, as the level of uncertainty or randomness in the data is quantified and more importance is placed on reliable and consistent criteria in the final decision [20].

The most attractive feature of this approach is the data-driven approach. Entropy weighting involves the computation of weights based on the data, as opposed to subjective weighting schemes, which make use of a subjective assessment or arbitrary decisions. This finds it especially applicable in situations when the expert knowledge is incomplete or absent. It is also natural in the ability of the method to normalize data, that is, the weights add to one. This fact simplifies the interpretation and comparison of the results. Moreover, the objective weighing technique is very adaptable, and it may be used in any range of subjects like economics, engineering, environmental science, and multi-criteria decision-making [21].

Moreover, this technique can be utilized to improve decision-making by considering the variability of the data [22]. This leads to more reliable and trustworthy outcomes while facilitating the balancing of different factors, ensuring that no single issue or factor can rule the decision-making process. This can be implemented because the weightages are based on data rather than personal bias. In short, the objective weighing method is a pertinent tool for both researchers and practitioners. The method has the capacity to improve decision-making quality due to its impartiality, ability to handle uncertainty, data-driven nature, and flexible. Consistent and clearer decisions can be made for different projects or technology choices by applying this tool, as it helps manage vagueness and offers comparing options.

2.4. Parameters Prioritization and Rank Parameters Based on Outcome

Prioritizing techno-economic parameters entails examining the most significant factors that influence the performance and viability of a technology or system, and concentrating the analysis on these factors. This ensures that limited resources are concentrated on the most critical issues and that the analysis is concentrated on the most critical information. The second phase of the research entails prioritizing the parameters derived in the first phase, utilizing the MCDM approach. The initial objective of this phase is to evaluate the significance of the parameters prior to prioritization. The MCDM method is regarded as an effective method for aiding decision-making under varying scenarios.

A common MCDM approach is as follows [23]:

Step 1: Identify the type of the problem.

Step 2: Construct a hierarchy of its evaluation.

Step 3: Select the most appropriate model of evaluation.

Step 4: Analyze the relative performance scores and the weights of each attribute relative to each alternative.

Step 5: Pick the best alternative based on the synthetic expected utility, which is the total of relative weights and performance ratings given to the alternatives. When the sum scores of the options are not definite, it will be helpful to insert Step 6 to establish a ranking and thereby make selecting the best alternative straightforward.

Step 6: Compare the alternatives on the basis of the synthetic values derived using fuzzy logic in Step 5.

2.5. Technique for Order Preference by Similarity to Ideal Solution

The Technique for Order Preference by Similarity to Ideal Solution is a ranking method that selects the most preferred alternative based on its proximity to the positive ideal solution and its distance from the negative ideal solution simultaneously [24]. A positive ideal solution refers to a solution that maximizes the desirable features and minimizes the undesirable attributes, and vice versa [22]. The TOPSIS technique offers four significant advantages, which are as follows: The logic of TOPSIS is rational and understandable. The calculation methods are straightforward, and the concept enables the identification of the best alternatives for each criterion using a simple mathematical representation. Additionally, the comparison operations incorporate the consideration of importance weights [25]. Regrettably, the TOPSIS technique has a primary drawback in that it incorporates a ranking index that takes into account distances from the ideal and negative-ideal points, but fails to examine their relative importance [26]. However, [27] adhered to the original approach, which includes the six successive steps of the TOPSIS approach operations as outlined by [28] as depicted in Fig. (1).

3.1. Parameters Prioritization and Ranking

A number of important parameters comprise the total investment cost per tonnage of production, the annual cost of raw materials per tonnage of production, the annual cost of utilities per tonnage of production, the annual fixed cost per tonnage of production, and the total revenue per tonnage of production are collected from six various processes of producing propylene oxide. Prior to solving MCDM problems, it is imperative to determine the weighted average of each technology route comparatively from each other. For the determination of the weighted average, the entropy method was selected because it is purely data-driven. It assigns weights based on the variability and uncertainty within the data itself, reducing the influence of subjective judgments [29]. In this study, the focus is on ranking the techno-economic parameters instead of the technology routes to see which parameter is the most important across the six different ways to produce propylene oxide. For a fair comparison, all parameters were calculated based on one tonne of production.

3.2. Entropy Method to Determine Weighted Average

The entropy method is an objective weighting approach widely applied in multi-criteria decision-making, performance evaluation, and composite index construction. It assigns weights to criteria based on their degree of variability (information content) within the dataset. Total Revenue is treated as a benefit criterion, where higher values are preferred. In contrast, Investment Cost, Raw Material Cost, Utilities Cost, and Fixed Cost are treated as cost criteria, where lower values are preferred. The procedure contains three core steps:

Step 1: Normalized the decision matrix with the help of Eq. (1),

Step 2: Compute the entropy using Eq. (2),

Step 3: Calculate the weight vector using Eq. (3),

3.3. TOPSIS to Prioritize and Rank the Parameters

In this study, the TOPSIS method will be used to evaluate and rank alternatives based on their closeness to an ideal solution. The principle is that the best alternative should have the shortest distance from the Positive Ideal Solution (PIS), representing the most desirable performance, and the farthest distance from the Negative Ideal Solution (NIS), representing the least desirable performance. The procedure involves seven main steps, as follows:

Step 1: Calculate Normalized Matrix using Eq. (4),

Step 2: Calculate the weighted Normalized Matrix using Eq. (5),

Step 3: Determine positive ideal solution with Eq. (6), negative ideal solutions with Eq. (7),

Step 4: Calculate the Euclidean distance from the ideal best using Eq. (8),

Step 5: Calculate the Euclidean distance from the ideal best using Eq. (9),

Step 6: Calculate performance score using Eq. (10),

Step 7: Ranking the performance score

4.1. Prioritization and Ranking Results from the TOPSIS Method

Table 5 indicates the weighted normalized parameter matrix per production, while Table 6 consists of a parameter matrix that has been weighted and normalized for each ton of production. Table 7 contains both positive and negative ideal solutions, while Table 8 indicates performance score results and techno-economic parameter ranking. The detailed TOPSIS calculation for the highest-ranked parameter, Total Revenue, is provided in Appendix B.

4.2. Robustness Analysis Using Monte Carlo Simulation

A Monte Carlo simulation was performed to evaluate the robustness of the Entropy-TOPSIS ranking, which involved all the original input parameters with ±10% variations. This uncertainty range was applied to reflect potential uncertainty in the techno-economic estimates, such as investment cost, total revenue, raw material cost, utilities cost, and fixed cost. The Entropy-TOPSIS procedure was repeated for each iteration of the simulation, and the ranking was compared with that of the base case ranking.

The results indicate that the total revenue and raw material cost remained in the top two parameters in 100% of simulation runs. This corroborates the major conclusion of the study, which is that total revenue and raw materials cost are relatively strong influences on the techno-economic parameters. Utilities cost, and investment cost show some sensitivity in the middle position, indicating that there is some interchangeability under uncertain conditions. Nevertheless, this finding does not impact the key findings of the study (Table 9).