Operational process and problems
Understanding Skinner (1969) in the sense that operational processes should be consistent with strategic goals, also considering Blenko et.al (2010) that the quality of decisions affects the success of a business, and Slack et.al (2016) and Heizer et.al (2017) studies that each process can be considered as an operational (transformation) process, where transformed resources using transforming resources through the transformation process give out a product or service that brings value to customers, and also, considering that each process can have sub-processes (process levels), which in turn can be associated with other processes; for this study, the gas transportation process is presented in Fig. 1 in a simplified form.
The objective of the gas transportation is to make profit by transporting customer’s gas from start point to end point. Thus, the transformed resources are gas, which, as a result of the transformation process (gas transportation process), transmitted from the start point to the end point. Also, gas suppliers and consumers can be considered as transformed resources, since as a result of the transformation the supplier gets rid of gas for a fee, and the consumer purchases gas, and accordingly their state changes due to transformation. Transforming resources are technical means (gas pipelines, compressors, etc.), pipeline operator and its employees (including contractors’ employees), its financial and other means, using which the transformation is performed.
Gas transportation process consists from many sub-processes, so within the framework of this work one of the most important sub-processes, namely the basic designing process (highlighted in yellow in Fig. 1) was analyzed. In this sub-process decisions on the main basic technical, economical, socio- ecological and other parameters of the project are determined and made.
In the process of basic designing, the main basic technical parameters such as the route and the length, diameter of the gas pipeline, required compressor powers and configuration are determined. To determine these basic parameters, need data, including on the current gas parameters (gas volume and pressure) at the inlet points and the required gas parameters at the outlet points, the location of the gas inlet and outlet points.
The sources of the data could be participants of the gas transportation process, e.g. gas suppliers and consumers, i.e. consumers can provide data on the gas demand, the locations to where the gas is to be delivered, its required parameters (volume, pressure), and gas suppliers providing data on location of existing or potential gas supply points and its parameters.
In more detail, the process of basic designing and use of the specified data could be described as follows: considering the location of the start and end points, site surveys of the potential routes will be carried out, based on the results of which the route and accordingly the length of the gas pipeline will be determined. Also, considering the required gas parameters at the end-outlet points and possible parameters at the start-inlet points, the required capacity of the gas pipeline, including the diameter of the pipes, power and configuration of compressors will be determined.
Fig. 1. Gas transportation process map.
The transportation process is associated with uncertainties and risks. One of the main uncertainties is the forecast of the gas supply and demand due to an open (indeterminate) future. According to the terminology of Hacking (1975) and considering the principles of Fox and Ulkumen (2011), these uncertainties can be attributed to the aleatory type. There may also be epistemic uncertainties, such as the lack of accurate data on sites, which will reduce the consideration of more alternatives to the
gas pipeline route; the lack of knowledge about the latest inventions, e.g. new pipes from stronger steels, which can reduce the consideration and selection of the best possible option for the thickness and, accordingly, the cost of pipes.
Considering Hillson (2018) about the differentiating risks with their causes and consequences, the following examples of decision risks are provided for the considered process:
- Impossibility of performing an accurate calculation of the basic parameters (diameter, powers, etc.) due to incompetent specialists (Risk 1).
- Impossibility to consider all alternative routes due to possible resistance of the local population and activists due to fear of threats to the environment and life (Risk 2).
- Accelerating and reducing the cost of the process due to a decrease or exclusion of site surveys in the event of receipt of data (topography, geology, etc.) from government or other data holders (Risk 3).
Decisions can be influenced by different biases as a result of the heuristics (Hardman and Harries, 2002). In the considered process, the following biases can be identified:
| Bias | Influences on the process decisions |
| Anchoring | Anchor data on gas supply and demand from certain provided or indicated data (often from the first indicated/provided). |
| Ambiguity or survivorship | Do not consider possible worst-case scenarios for the gas market development, for which favorable business development is not known, e.g. a sharp decline in demand due to the development of new technologies. |
| Confirmation | Seek and use only information, such as the growth in gas demand and supply/production, that confirms established belief that the market will grow. |
| Availability | Consider only those technologies (e.g. pipe materials, compressor drives) for which can recall from previous experience or the prevalence of these technologies. |
Also, the decisions made could be influenced by the framing effect. The process steps and final decisions can be influenced by decisions made by external stakeholders, e.g. activists or the public can disrupt performing site surveys to explore and select the optimal pipeline route. The following examples can be provided for the three types of framing mentioned by Levin et al. (1999):
| Type | Example and affect to process |
| Risk Choice Framing | Depending on how the safety of the gas pipeline for surroundings will be indicated (e.g. ‘it is 99% safe at a distance of 300m and beyond’ or ‘1% of the probability of the pipeline explosion’), activists or surroundings can make a positive (in the first example) or negative (in the second example) decision. |
| Attribute Framing | The decision can be made in favor of a new type of gas turbine engine for which the best attributes are highlighted (e.g. it is indicated that it is more efficient), but in fact this new type may be less reliable (prone to frequent failures due to the complicated design) and more expensive to maintain. |
| Goal Framing | If it will be indicated that if the new gas pipeline will be not built the population cannot be provided with a enough gas, it may lead to a positive decision by the population and activists regarding the project implementation. |
Also, some participants (e.g. gas suppliers or consumers) can manipulate data, e.g. indicate large supplies or consumption in order to book large capacities. This lack of data quality can lead to wrong decisions, e.g. to excessive capacity and, accordingly, to unnecessary capex.
Recommendations for improvement
As found above, the considered process can be susceptible to poor decision making due to poor data quality, due to a person’s adherence to heuristics leading to biases in decision making, due to framing of the issues. This section discusses and recommends measures to improve the quality of decisions.
Considering that data is foundation of Thieraufs (1999) data-information-knowledge hierarchy, and as noted by O’Reilly (1982) and later by de Fortuny et al. (2014), Sadiq and Indulska (2017), Janssen et al. (2017) that data quality is important, in order to prevent deception and incorrectness of the converted information, the first to improve the quality of decisions can be recommended to improve the process of converting data to information and to knowledge using wisdom.
As mentioned above, some stakeholders (gas suppliers and consumers) can manipulate data in their own interests, and considering that the data mining process can help to identify patterns and knowledge (Han et al., 2012), the data mining process example to increase the accuracy of the future gas demand is indicated in Fig. 2.
o Using AI technologies such as Data Science and Neural Networks, determine the interrelationship of the trend in the consumption of natural gas and other energy sources.
o By performing regression analysis, including using AI technologies, to predict the demand for natural gas in the future. For example, this interrelationship can be as shown in Fig. 3.
s Based on the results of these two studies (descriptive and predictive), using wisdom, determine the most likely forecasts of gas demand.
Considering the above probabilities of data manipulation, the cost of making bad decisions to the business, the presence of a large amount of data, including unreliable ones, it is proposed and is valuable for this process to use data scientists. AI technologies such as machine learning and neural networks can be used to support data scientists in processing unstructured data.
To prevent Attribute Framing affect noted above (e.g. a new model of a gas turbine engine or an existing tested model), when there is accurate data (e.g. technical parameters for engines are known that can be compared) can follow prescriptive approaches and apply decision aids, e.g. multi-criteria technique. An example of using the Simple, Multi-Attribute Rating Technique (SMART) proposed by Edwards (1971) for the indicated problem of choosing between different engine models is carried out as provided below.
| Table 1. Multi-Attribute Utility (MAU) Calculation. |
| Table 2. Costs calculation for 10 years of operation. Fuel gas consumption Overhaul Capital Total cost Parameter 9245 Btu/kWh 40k hrs 24.28 Gt IM2500 Cost 8.73 2.55 13 GE LM2500 Parameter 9352 Btu/kWh 50k hrs 22.27 + G4 Cost 8.83 2.04 11.4 R-R RB211- Parameter 8956 Btu/kWh 30k hrs 22.06 300 Cost 8.46 3.4 10.2 Siemens Parameter 9150 Btu/kWh 40k hrs 22.69 SGT-700 Cost 8.64 2.55 11.5 Note: Costs are in MM USD |
Example of SMART analysis sequence:
- By exploring the market, from experience of previous projects and information from other similar projects, were selected a list of alternative models of gas turbine engines that correspond to the required engine power (calculated by special software in advance, e.g. 30 MW), and not violate the requirements for emissions, noise, etc.
- Identified the most important attributes, including natural and judgmental, for comparison and scoring. Technical knowledge and experience were used for scoring the attributes (e.g. more power, less emission and less noise is better), and for judgmental attributes subjective assessment.
- Depending on the importance of the attribute, considering the technical knowledge and experience, weightings were set, and normalized weightings were calculated.
- Considering the normalized weightings (attribute importance), the MAU of each alternative was calculated.
- To calculate the cost, capex and opex were considered.
- Opex were estimated considering the technical parameters, e.g. gas consumption of the engine at the same power of 30 MW and 8.5 thousand operating hours per year for 10 years, and the cost of this gas. The cost of maintenance was calculated considering the number of overhauls within 10 years and the average cost of one overhaul.
- Next, considering the capex of the engine, the total cost of each engine for 10 years was calculated.
Next, the MAU and total cost indicators were displayed in a chart to visualize the attribute value and cost, as shown in Fig. 4.
Fig. 4. MAU vs cost chart.
SMART analysis can help to choose the best option from the alternatives, in this example it would be GE LM2500+G4.
Gas pipeline projects require huge investments and, accordingly, operators strive to ensure the sustainability of their business and reduce the risks associated with uncertainties. In such cases, operators usually seek to enter into long-term contracts with consumers for the transportation of a certain volume of gas with a ‘transport or pay’ terms, which means that even if the consumer nominate for smaller volumes he will have to pay for the volumes specified in the agreement. The Decision Analytic approach suggested by Raiffa (2001) could be useful for negotiating this kind of agreements in order to make it beneficial to both parties. In this approach the parties must prepare for negotiations, determine the interests of both, their importance, reservation points, and this could look like this:
| Information | Operator | Consumer |
| BATNA | From the beginning it is required that through data mining and other methods, as accurately as possible, determine the possible future volumes of gas demand by consumers. This will be an alternative for the operator if an agreement fails, i.e. the operator can design the pipeline based on the most accurate knowledge he could determine. | If the consumer does not conclude an agreement and the operator refuses to construct a gas pipeline without it, the alternative for consumer will be to use other energy sources that are next in attractiveness (LNG, hydrogen, etc.). |
| Reservation point | Based on the above knowledge, the operator will be able to assume and establish a reservation point, both his own and the consumer’s, in this example the expected volume of transportation within a certain period. Another reservation point for the operator will be the duration of the agreement. This can be determined based on how long it will take for the project to pay off with the considered gas transportation volumes. | For a consumer a reservation point can be the volume and the duration of the agreement when losses due to the possible lower consumption will be less than the possible losses due to the higher cost of other energy sources. |
| Information | Operator | Consumer |
| Interests and order of priority | 1. Agree on the minimum volume that would guarantee the payback of the project. 2. Determine equal transportation volumes throughout the year. | Get access to pipeline gas.Assure the appropriate volume, considering the foreseen consumption. |
Possible risks (negative and positive) were identified above. Below are examples of measures to eliminate or reduce negative ones and increase the possibility of positive ones.
| Risk | Consequences | Measures |
| Risk 1 | Negative | Develop criteria for selecting engineering companies and their employees, e.g. experience, education, etc. |
| Risk 2 | Negative | Elucidate the safety of the gas pipeline (calculations, used computing tools), show the positive aspects of the project, such as the creation of new jobs. |
| Risk 3 | Positive | Conclude non-disclosure agreement, interest government agencies in the benefits of gasification, e.g. reducing CO2 emissions. |
Since the considered process is in fact a design of the business future, process decisions must consider the triple bottom of Elkington (1994) for business sustainability. Examples of such considerations can be making decisions to consider the opinions of external stakeholders (social, etc.) on the project, choosing eco-friendly solutions (more efficient compressors, applying recovery of exhaust gases technologies, etc.).
Given that operational (transformational) processes can have an impact on strategic and social objectives, operational performance objectives can be addressed according to the following 5 metrics identified by Slack et al. (2016):
Quality
Quality standards (procedures, policies) shall be established, e.g. follow the ISO 9000 series standards to ensure high quality designing and thus ensuring high quality the first time.
Speed
It can be established requirements for the number of personnel, for the use of modern design technologies (3D, etc.) which will ensure execution on time.
Dependability
To control the quality and speed of the process, the operator must also have enough abilities, e.g. have qualified specialists, have the financial capabilities (cash flows, etc.) to timely finance the process.
Flexibility
Operator and engineering company must react quickly and adapt to changes, e.g. to actions of population, activists; to the emergence of new products (pipe materials, compressor models).
Cost
To seek opportunities to reduce costs, e.g. obtaining survey data from government; also, the decisions of the process should ensure the profitability of the business in the future.
The following Intelligent DSS technologies can be applied and assist in decision making during the following stages of the process:
- As noted above, there is a risk of prevention by activists and local communities to perform several steps of the process. In this case, along with the above communication measures, it can be observed and studied the views of the population and propaganda carried out by activists by processing big data, e.g. from social networks, news portals.
- Various expert systems, e.g. special optimization software to choose pipeline route, compressor configuration, pipe materials can be used. It is recommended to consider the results of these systems as supportive in decision making, since the algorithms may contain some biases, e.g. restrictions for the compressor configuration by power.
- It is also useful to conduct a project premortem to identify possible improvements to the project, considering the views of various stakeholders (scientists, population, etc.). In this case, the population can point some problem areas and places, scientists can suggest new technologies that were not known to the operator and engineering companies.
Was developed the process map, from which it is easier to determine the key decision points, which in turn will lead to a better understanding of the process, its weaknesses, and to provide measures to improve the decision making.
The analysis revealed that biases and framings can affect decision-making, and examples of such biases and framings are provided. In order to avoid making the bad decisions because of this, and as far as possible to eliminate the influence of biases and framing, it is recommended to follow a prescriptive approach, using structured decision aids. At the same time, it is worth noting that in some cases heuristics can be useful, e.g. the experience and knowledge of experts to make quick decisions in cases of uncertainties.
Recommendations for using the capabilities of modern technologies such as big data, AI are given. That said, it is worth considering hype with big data, and the need to use data scientists to generate useful information (Provost and Fawcett, 2013). It was also noted that given the modern limitations of AI, it is recommended to use them as supporting systems (Augmented Intelligence) in decision making.
Note: Incorrect Referencing.
- Skinner, W. 1969. Manufacturing-missing link in corporate strategy. Harvard Business Review, 47(3), pp.136-145.
- Blenko, M., Mankins, M., and Rogers, P. 2010. The decision-driven organisation. Harvard Business Review, 88(6), pp.54-62
- Heizer, J., Render, B., and Munson, C. 2017. Operations Management: Sustainability and Supply Chain Management (Global edition). Pearson Education.
- Slack, N., Brandon-Jones, A. and Johnston, R. 2016. Operations Management. Pearson Education.
- Thierauf, R. J. 1999. Knowledge management systems for business. Greenwood Publishing Group.
- Hacking, I. 1975. The Emergence of Probability. Cambridge University Press.
- Fox, C. R., and Ulkumen, G. 2011. Distinguishing two dimensions of uncertainty. In: Brun,W., Keren, G., Kirkeb0en, G. & Montgomery, H. Perspectives on thinking, judging, and decision making, (eds) Universitetsforlaget, Oslo.
- Hillson, D. 2018. When is a Risk not a Risk? IACCM.com. [Online]. 29 May [Accessed 30 October 2020].
- Hardman, D. and Harries, C. 2002. How rational are we? The Psychologist. 15(2), pp 76-79.
- Edwards. W. 1971. Social utilities. The Engineering Economist, 6, pp.119-129.
- Raiffa, H. 2001. Collaborative decision making. Cambridge, MA: Belknap.
- Provost, F., and Fawcett, T. 2013. Data science and its relationship to big data and data-driven decision making. Big data, 1(1), pp.51-59.
- Skinner, W. 1969. Manufacturing-missing link in corporate strategy. Harvard Business Review, 47(3), pp.136-145
- Slack, N., Brandon-Jones, A. and Johnston, R. 2016. Operations Management. Pearson Education.
