Process Intensification and Integration for Sustainable Design. Группа авторов

rate and composition pose major challenges when designing a gas processing plant of optimal size. In general, plants with larger process equipment are more flexible and are able to handle a wider range of inlet compositions. Nonetheless, these plants also have higher fixed and variable costs. A gas processing plant is needed to purify and separate natural gas and natural gas liquids (NGLs) and to isolate various possible containments including water, sulfur species, carbon dioxide, mercury, and oxygen [6]. Such separation operations may include acid gas removal, to remove sulfur species and carbon dioxide, dehydration, nitrogen rejection, mercury removal, NGL recovery, and NGL separation. One issue currently facing the gas production industry is a lack of capacity to handle greatly increased production [7]. Another issue is frequent unplanned shutdowns and a lack of efficiency in operations [8]. Regardless of the dynamic and spatial variability in shale gas flow rate and composition, gas processing facilities must have the ability to handle such variations and render a set of products with consistent qualities to satisfy pipeline constraints and downstream‐processing requirements [9-12]. In this chapter, the aim is to determine a method to find the optimal size of a plant and a strategy to process wellhead gas when feeds of various compositions are available to the facility. Process synthesis, simulation, and techno‐economic analysis were used to determine the optimal configuration and capacity of the gas treatment plant.

      The approach will also incorporate safety into the early stages of process design, before changes in design become more costly and difficult to make [13-15]. The concept of inherent safety is that, by eliminating or reducing the sources of hazards in a chemical plant, the severity and likelihood of process safety incidents will be reduced [8]. One challenge of implementing inherent safety is the lack of information in early design stages. Most existing safety assessment tools are used retroactively, after the process design is completed or near completion [16]. In order to quantify the inherent safety of alternative process designs during the early design stages, a number of safety indices have been developed [14,15,17]. In this work the safety of different process designs will be compared using a modified version of the process route index (PRI) [18]. This safety index was chosen because the chemicals involved in natural gas processing are highly flammable and explosive [18,19].

      Another important consideration is environmental impact. While natural gas is considered to be cleaner than coal and oil (from an emissions and energy consumption standpoint), there is potential for further reduction in environmental impact [20,21]. However to the author's knowledge fluctuating feedstock compositions have not been considered in literature for shale gas processing.

2.2 Problem Statement

      The problem to be addressed in this work is stated as follows:

       A set of shale gas wells with anticipated profiles for variable flow rates and compositions and known, temperature, and pressure

       A known set of feedstock and product prices

It is desired to develop a systematic design and optimization approach for gas treatment plant. Although the processing steps for treating raw shale gas can vary depending on the composition of the wellhead gas, a common process flowsheet (shown by Figure 2.1) is considered. First, condensates and free water are separated. Then, acid gases (CO₂ and H₂S) are removed. Acid gas content must be lowered to permissible levels to prevent corrosion issues during additional processing and/or during pipeline transport [6]. Dehydration is then carried out to remove bound water. This water must be removed to low levels because (i) it may form hydrates with natural gas components such as methane, ethane, and carbon dioxide, and (ii) it may freeze in later processing steps. In either case, the formed solids may plug piping and separation units [6]. Next, methane (sales gas) is separated from the other NGLs. Finally the latter is fractionated into their individual components as they typically have high economic value.

Figure 2.1 Shale gas treatment process.

2.3 Methodology

Figure 2.2 gives an overview of the proposed approach. First, a statistical analysis is carried out for the composition data. Several feeds are chosen, and process simulations are performed using a set of given assumptions to meet product specifications. Then, a process design is developed for each feed. Next, fixed and variable costs are estimated using simulation results, which include equipment sizing, mass and energy balances, operating conditions, and utility consumption as well as detailed cost data. Finally economic calculations are performed, and when combined with revenue information enable analysis of economic results.

Figure 2.2 Overview of the methodology.

2.4 Case Study

      To illustrate the applicability of the proposed approach, a case study is solved based on representative data for the Barnett Shale Play in Texas. The key objectives of the case study include:

       Design of a base case and several additional process designs for different feed compositions

       Economic evaluations of the proposed designs

       Process safety evaluation of the proposed designs

       Sensitivity analysis where product and feedstock prices are varied based on standard deviations from historical price data

      To streamline the study, the following assumptions are made:

       Average flow rate, temperature, and pressure: Although the flow rate, temperature, and pressure of shale gas coming out of the well can vary significantly, it was assumed that wellhead gas is sent to a centralized processing facility where these values on average would be relatively constant, and only composition would vary. Additionally it is common for gas to be saturated with water because water is sent down the well to maintain well pressure. It is not uncommon for there to also be free water in the incoming gas stream; however this is easily removed using a knockout drum on the front end of the process at minimal cost.

       Inlet feeds enter the processing plant at a standard vapor volumetric flow of 150 million standard cubic feet per day (MMSCFD), 100 °F, and 1000 psig. The Peng–Robinson equation of state was used in the process simulation model.

       The gas feedstock is saturated with water.

      2.4.1 Data

First, compositional data was obtained for the Barnett Shale region (located in the Dallas, TX, area) [3]. Next, minor components such as nitrogen, oxygen, and hydrogen were removed from the data sets so that only the major components (CO₂ and the hydrocarbons) were considered. The compositional data sets were classified into different types. This was done based on the methane composition in each data set. There were five types of data sets as shown in Table 2.1. Table 2.1 also shows the probability of each feed type as determined from the literature data [3].

Table 2.1 Feed types as determined by methane composition.

Feed type	Methane composition (mol%)	Probability of feed type (%)
1