(371n) A Novel Methodology to Assess Safety of Process Streams at the Conceptual Design Stage

The traditional and most applied way to design a process considers the fulfilment of technical and economic targets as the main objectives, while environmental and safety matters are often addressed after the main decisions have been made, with limited room for improvement. This not only undermines the optimal overall performance of the technology but also might conduct to further economic investments to adjust some parts of the project to meet safety or environmental requirements if lacks are individuated.

In this context, a research need emerges: considering a holistic vision during the conceptualization of a process.

Safety specifically must always come first among the design objectives. Along with being hazardous, an unsafe plant cannot be profitable too, due to losses of production and capital that an accident may cause. Given these reasons, safety considerations should shape design decisions from the initial stages of a design project, not after the detailed design phase is finished [1]. Within this traditional approach, there is limited opportunity to alter the design to improve safety performance. An alternative and beneficial approach is to integrate inherent safety principles during the chemical process design stage [2].

The assessment of inherent safety principles in early process design has been recognized as one of the major topics under current research. Particularly, over the past few decades, there has been a proliferation of inherent safety metrics for measuring, ranking, and selecting process alternatives.

Inherent safety metrics have been gaining success due to their straightforward implementation, minimal data requirements (making them suitable for the conceptual stage), and ability to provide immediate feedback on process safety performance.

These metrics can be classified into four categories: consequence-based metrics, graphical assessments, risk-based metrics, and index-based metrics. The index-based metrics are the most targeted because they consist of a mathematical model that produces a number as a single output, that ranges over a scale [4]. Even though an assessment done by a team of experts can never be replaced by an automated method, linking the assessment of safety to process simulation simplifies, systematizes, and speeds up the design process [8].

Although major advances have been observed in the development of safety indices, there is no unified metric for assessing inherent safety [3], [5], [6]. Hence, there is still a demand for new simple indices for use in the conceptual design phase [7].

The literature review we conducted highlighted that a simultaneously comprehensive, reliable and user-friendly index does not exist. Furthermore, many indexes lack a standard scale (e.g., from 1 to 10), making it impossible to understand the value of the index alone if no direct comparison with another alternative is provided.

Here we develop and implement an index capable of evaluating the safety of a generic process stream, comprehensive of all the possible information we have at the conceptual stage.

The index is developed as a MATLAB code coupled with Aspen Plus, that calculates heat and energy balances. The code receives the parameters computed by Aspen Plus as input, except for a few parameters that are provided manually.

The index gives immediate feedback about the safety of the stream, and when the process conditions change, the index is recalculated, giving immediate feedback on how different choices impact the performance.

We also conducted some sensitivity analyses to evaluate how a particular parameter impacts the value of the index, and to understand what its percentage variation is once some parameters are changed.

Here we start from the analysis of process streams because a simulator like Aspen Plus gives us a lot of properties about streams rather than units, and we believe that once you have the indexes for the various streams, it is straightforward to get the indexes for the various units.

Our objective is to have an index that is also easy to understand for the final user. For this purpose, we employed a simple scale from 1 to 10: the higher the number the better the index performance.

The code is organized into different sections, one for each property analyzed. We get one index for each section first, and then all the indexes are combined to obtain the index for the stream.

We selected the properties for the index calculation according to some literature reviews that have collected and classified the main properties necessary to assess inherent safety at conceptual stage [3], [4], [8].

Here we want to propose something different from what is presented by other scientific publications, where safety indexes are built according to penalties given by step functions [1], [9], [10]. The main limitation of this approach is that it sets a fixed penalty value for a wide range of values. For instance, we cannot possibly affirm that a vessel operated at 0.5 bar and at 5 bar is hazardous in the same way.

To avoid this, we produced several continuous functions with a shape that fits the trend of some assigned combinations of points.

All the functions are characterized by a fixed range on the y-axis, that goes from 0 to 10, as the range of the index. On the x-axis there are the values of the parameter under study.

The points are selected according either to the literature [1], [4], [6], [9], or to the opinion of some experts in the field, or the combination of both.

The functions we used are the logistic, the logarithmic, and the power. The functions are purposely intended to be parametric. For every function, an optimization code has been implemented into MATLAB. This, according to the fixed points and the initial guesses, retrieves the optimal values of the parameters needed to fit the desired function.

This procedure allows us to avoid all the step functions and ranges, and to have a neater code for the computation of the safety index.

The implementation of the methodology is critically tested on the streams of a case study flowsheet implemented into Aspen Plus. Results show that our methodology is capable of identifying the most hazardous streams within a process, allowing the designer to make different choices in order to improve the overall safety performance. Finally, we want to underline that this is only the starting point for a more comprehensive methodology, capable of assessing the inherent safety of all the units involved in a process.

References

[1] A. Heikkilä, “Inherent safety in process plant design, Technical research center of Finland,” pp. 1–132, 1999.

[2] M. Athar, A. M. Shariff, A. Buang, M. S. Shaikh, and T. L. See, “Inherent safety for sustainable process design of process piping at the preliminary design stage,” J Clean Prod, vol. 209, pp. 1307–1318, Feb. 2019, doi: 10.1016/j.jclepro.2018.10.281.

[3] X. Gao, A. A. Abdul Raman, H. F. Hizaddin, M. M. Bello, and A. Buthiyappan, “Review on the Inherently Safer Design for chemical processes: Past, present and future,” J Clean Prod, vol. 305, p. 127154, 2021, doi: 10.1016/j.jclepro.2021.127154.

[4] S. Park, S. Xu, W. Rogers, H. Pasman, and M. M. El-Halwagi, “Incorporating inherent safety during the conceptual process design stage: A literature review,” Journal of Loss Prevention in the Process Industries, vol. 63. Elsevier Ltd, Jan. 01, 2020. doi: 10.1016/j.jlp.2019.104040.

[5] J. Zhu, Z. Liu, Z. Cao, X. Han, L. Hao, and H. Wei, “Development of a general inherent safety assessment tool at early design stage of chemical process,” Process Safety and Environmental Protection, vol. 167, pp. 356–367, Nov. 2022, doi: 10.1016/j.psep.2022.09.004.

[6] Y. Qian, S. Vaddiraju, and F. Khan, “Inherent Process Risk Index (IPRI) – A tool for analyzing Inherently Safer Design using Aspen Plus simulation,” Process Safety and Environmental Protection, Jan. 2024, doi: 10.1016/j.psep.2023.12.070.

[7] M. H. Ordouei, M. Elsholkami, A. Elkamel, and E. Croiset, “New composite sustainability indices for the assessment of a chemical process in the conceptual design stage: Case study on hydrogenation plant,” J Clean Prod, vol. 124, pp. 132–141, Jun. 2016, doi: 10.1016/j.jclepro.2016.02.107.

[8] H. Mohammadi, M. J. Jafari, M. Pouyakian, E. Keighobadi, and S. Moradi Hanifi, “Development of a new index for assessing the inherent safety level of chemical processes using a multi-criteria fuzzy decision-making approach,” J Loss Prev Process Ind, p. 105238, Dec. 2023, doi: 10.1016/j.jlp.2023.105238.

[9] P. Gangadharan, R. Singh, F. Cheng, and H. H. Lou, “Novel methodology for inherent safety assessment in the process design stage,” Ind Eng Chem Res, vol. 52, no. 17, pp. 5921–5933, May 2013, doi: 10.1021/ie303163y.

[10] M. Athar, A. M. Shariff, A. Buang, A. Umer, and D. Zaini, “Inherently safer process route ranking index (ISPRRI) for sustainable process design,” J Loss Prev Process Ind, p. 104909, Dec. 2022, doi: 10.1016/j.jlp.2022.104909.