A graphical technique for wastewater minimisation in batch processes

Journal of Environmental Management 78 (2006) A graphical technique for wastewater minimisation in batch processes Thokozani Majozi a, *, C.J. Brouckaert b, C.A.
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Journal of Environmental Management 78 (2006) A graphical technique for wastewater minimisation in batch processes Thokozani Majozi a, *, C.J. Brouckaert b, C.A. Buckley b a Department of Chemical Engineering, University of Pretoria, Lynnwood Road, Pretoria 0002, South Africa b Department of Chemical Engineering, University of KwaZulu Natal, Durban 4001, South Africa Received 20 April 2004; revised 24 February 2005; accepted 14 April 2005 Available online 22 August 2005 Abstract Presented in this paper is a graphical technique for freshwater and wastewater minimisation in completely batch operations. Water minimisation is achieved through the exploitation of inter- and intra-process water reuse and recycle opportunities. In the context of this paper, a completely batch operation is one in which water reuse or recycle can only be effected either at the start or the end of the process. During the course of the operation, water reuse and recycle opportunities are completely nullified. The intrinsic two-dimensionally constrained nature of batch processes is taken into consideration. In the first instance, time dimension is taken as a primary constraint and concentration a secondary constraint. Subsequently, the priority of constraints is reversed so as to demonstrate the effect of the targeting procedure on the final design. Attention is brought to the fact that first and cyclic-state targeting are essential in completely batch operations. Moreover, the exploration and use of inherent storage in batch processes is demonstrated using a real-life case study. Like most graphical techniques, the presented methodology is limited to single contaminants. q 2005 Elsevier Ltd. All rights reserved. Keywords: Reuse; Recycle; Completely batch; Water minimisation 1. Introduction The last two decades have been characterised by intensified research in the area of mass integration in chemical processes. This is a direct result of the evertightening environmental constraints as well as increased global awareness on sustainable development, hence cleaner production. However, most of the developments have been targeted at continuous processes at steady-state (Takama et al., 1979; El-Halwagi and Manousiouthakis, 1990; Wang and Smith, 1994, 1995a,b; Kiperstok and Sharratt, 1995; Olesen and Polley, 1997; Jödicke et al. 2001; Hallale, 2002). Equal developments in batch chemical processes are still at their infantile stages for several reasons. Firstly, most mass integration techniques are founded on heat integration, which is widely known as pinch technology. It was initially thought that the application of pinch technology based techniques in batch processes would have very limited benefits due to the * Corresponding author. Tel.: C ; fax: C address: (T. Majozi) /$ - see front matter q 2005 Elsevier Ltd. All rights reserved. doi:116/j.jenvman intrinsic time constraints. In essence, it is mainly for this reason that most process integration methodologies in batch operations are based on mathematical programming, which is virtually dimensionally unconstrained, rather than on graphical analysis (Vaselanak et al., 1986; Grau et al., 1996; Sanmartí et al., 1998; Yao and Yuan, 2000; Majozi and Zhu, 2001). Secondly, the incorporation of process integration in batch processes was perceived to lead to reduced flexibility, which is the main feature of batch processes. Thirdly, there is a general understanding that batch processes have intrinsic variations, which eventually lead to deviations from predetermined schedules. This would make targeting extremely difficult, if not impossible. Fourthly, as pinch technology was regarded as only an energy optimisation technique, it was assumed that process integration would not have much significance in batch operations, since energy and water usually constitute a small component of operating costs (Obeng and Ashton, 1988). Stringent environmental regulations are steadily rendering this notion untrue. Recently, Foo et al. (2004) have developed a methodology for the synthesis of mass exchange network in batch processes with a focus on utility or mass separating agent (MSA) targeting. This methodology is an adaptation of 318 T. Majozi et al. / Journal of Environmental Management 78 (2006) the work by El-Halwagi and Manousiouthakis (1990) for mass exchanger network synthesis in continuous processes, combined with the cascade analysis for batch heat integration developed by Kemp and Deakin (1989). A new targeting tool called time-dependent composition interval table that facilitates vertical cascading of mass exchange between rich and lean streams through composition intervals forms the basis of this methodology. The main drawback of this technique is its limitedness to operations that involve mass exchange between lean (solvent) and rich (process) streams for which equilibrium data readily exists. In a situation where the removal of mass load does not entail mass transfer, e.g. washing operations, this procedure cannot be directly applied. Certainly, in order to handle batch operations effectively, the time dimension cannot be ignored due to the fact that almost all operations within the batch process environment are time dependent. Fig. 1 shows comparison between batch and continuous processes on the exploration of water reuse opportunities. In continuous operations, only the concentration constraint determines the feasibility of water reuse from one process to another. This implies that if the outlet water concentration from process A is less than the maximum allowed inlet water concentration to process B, then water from process A could be reused in process B. On the other hand, if water from one process, say process B, is at a concentration higher than the maximum allowed in another process, say process A, the water reuse opportunity from process B to process A is nullified. In batch operations, there is also a time constraint to be satisfied, in addition to the concentration constraint. This particular feature renders batch operations more challenging than their continuous counterparts. As shown in Fig. 1, even if the concentration constraint is obeyed, water from process A cannot be readily reused in process B if process A commences after process B. This constraint could be bypassed by using reusable water storage tanks wherein water is stored for later use. However, this cannot be efficiently achieved without properly addressing the time constraint that is inherent in batch operations. In 1995, Wang and Smith developed a landmark graphical analysis technique for water minimisation in Continuous processes vs Batch processes Water reuse Water reuse time A A concentration A A B B concentration Fig. 1. Comparison of continuous and batch processes. B B batch processes, which forms the basis for the argument put forth in this paper. A detailed presentation of this technique is, therefore, mandatory in order to decipher the subsequent sections of this paper Wang and Smith approach for water minimisation in batch processes To facilitate understanding of their methodology as applied to setting water demand and wastewater generation targets, Wang and Smith (1995a,b) formulated an example comprising of three processes as shown in Table 1. It should be noted that these data are limiting, i.e. the inlet and outlet concentrations of contaminant have been set to the maximum values. Ignoring time constraints and solving the problem as a continuous process yields water demand target of 155 t/h. However, the fact that this is a system of batch operations means that the inherent time constraints cannot be ignored. The implication of the time constraint in batch operations has been presented earlier, Fig. 1. Since mass transfer takes place within a limited period of time, which is not the case with continuous processes, the amount of contaminant transferred is measured in kg instead of kg h K1. The amount of contaminant is calculated as: Dm Z f DCDt (1) where DC is the change in concentration, Dt is the duration of mass transfer and f is the water flowrate required. To understand the role of the time dimension better, Eq. (1) is rewritten as: Dm Z f Dt (2) DC In Eq. (2), fdt is the amount of water required to remove Dm (kg) of contaminant associated with a DC change in concentration. If DG is assigned to the amount of water required, the following equation is obtained: DG Z f Dt (3) Plotting DG against time as shown in Fig. 2 clearly represents the behaviour of a batch water-using operation. To ensure that both concentration and time constraints are met, this analysis should be applied in each of the concentration intervals. It would be expected that a similar approach to the Problem Table Algorithm (Linnhoff and Flower, 1978) should be applied in order to ensure that a specific minimum concentration difference holds in each of Table 1 Data for the Wang and Smith example Process number Flowrate (t/h) s (ppm) C in,max C out,max t 0 t T. Majozi et al. / Journal of Environmental Management 78 (2006) Water required ( G) G (t) G 1 Process 2 40 G 0 the intervals. This would entail shifting of the inlet and outlet concentrations for the process streams as it was done in setting the energy targets. However, the fact that the limiting concentration constraints have been built into the problem makes the shifting of concentrations irrelevant. The concentration intervals are demarcated by the inlet and outlet concentrations as shown in Fig. 3. Targeting is performed by cascading water from one concentration interval to the next, until the last concentration interval. Within each concentration interval, water is cascaded from one time subinterval to the next without degradation. This appears to be the mass transfer version of the targeting procedure presented for heat exchangers. In concentration interval (0 100 ppm) only process 2 exists, Fig. 3. Targeting in this concentration interval is shown in Fig. 4. The amount of water required is 40 t, which leaves the process at a concentration of 100 ppm. In the concentration interval ( ppm), processes 1, 2 and 3 coexist as depicted in Fig. 3. The total amount of water required is: ½100ð1:5 K0:5ÞŠ process 1 C½80ð0:5 K0ÞŠ process 2 C½50ð1:0 K0:5ÞŠ process 3 Z 165 t (4) This water should be supplied at a concentration of 100 ppm. Targeting in this concentration interval is shown in Fig. 5. However, only 40 t of water is available from the first concentration interval. This implies that the deficit should be supplied by fresh water. The amount of fresh water (ppm) [2] t 0 t 1 [3] [1] Fig. 3. Demarcation of concentration intervals. Time Fig. 2. A batch water-using process on water required versus time. Fresh water required is calculated as: ð165 K40Þð200 K100Þ ð200 K0Þ 0.5 Fig. 4. Targeting in concentration interval (0/100 ppm). Z 62:5 t (5) Therefore, the total amount of contaminated water is now t at a concentration of 200 ppm. This water can be reused in the next concentration interval. In concentration interval ( ppm), only process 1 exists, Fig. 3. Targeting in this concentration interval is shown in Fig. 6. The total amount of water required is 100 t at a concentration of 200 ppm. Since t of water at a concentration of 200 ppm is available for reuse from the previous interval, no fresh water is required in this interval. This eventually sets 200 ppm as the pinch concentration, i.e. the concentration beyond which the water target does not change. The total amount of wastewater generated from processes is t, which is the minimum wastewater target. G (t) Freshwater 62.5 tons Composite of the processes Fig. 5. Targeting in concentration interval (100/200 ppm). G (t) Water available for reuse Storage Composite of the processes Fig. 6. Targeting in concentration interval (200/400 ppm). 320 T. Majozi et al. / Journal of Environmental Management 78 (2006) As it can be inferred from Fig. 6, in the time period (0 0.5 h), 40 t of water is available for reuse. In the time period ( h), 50 t of water is required, but there is only 37.5 t of water available. Therefore, 12.5 t of water is required from storage. In the time period ( h), 50 t of water is required and only 25 t is available, thus 25 t of water is required from storage. This implies that 37.5 t of water from time period (0 0.5 h) should be stored for reuse in the time periods ( h) and ( h). Fig. 7 shows the design to meet the target. Implicit in their analysis (Wang and Smith, 1995a,b), were a number of assumptions which were not immediately obvious. These were surfaced when this methodology had to be applied to a set of completely batch processes in an agrochemical facility. Firstly, the analysis implicitly allows water reuse even in a situation where the source and sink processes are simultaneously active as observed in Fig. 6. According to Fig. 6, in concentration interval (200 to 400 ppm), 37.5 and 25 t of water are available for reuse in the (0.5 to 1.0 h) and (1.0 to 1.5 h) time intervals, respectively. This semi-batch behaviour is further observed in the water network shown in Fig. 7. Some of the water from process 3 is reused in process 1, even though process 1 is 0.5 h from completion. This would not be possible for truly batch operations, since the reuse potential of water could only be realized after the completion, and not during the course of the process. Secondly, the network layout showed in Fig. 7 shows that 12.5 t of water should be supplied to Process 3, instead of 25 t stipulated in the problem specification. This can only be true if this process does not have flowrate constraints, but has a fixed mass load. The assumption of fixed mass loads was never mentioned in the analysis. This variation of flowrate is contrary to the assumption made in targeting. During targeting it was implicitly assumed that the flowrates were fixed as shown by the calculation of water demand in each of the time subintervals t Thirdly, the inlet and outlet concentrations were specified such that one was fixed directly and the other determined by mass balance using flowrate and mass load. However, a number of variations are possible in the way that the process constraints on quantity (or flowrate) present themselves. For instance, it could happen that there is no direct specification of the water quantity (or flow) in a particular stream, as long as the contaminant load and the outlet concentration are observed. Furthermore, the vessel probably has minimum and maximum levels for effective operation. In that case the water quantity falls away as an equality constraint, to become an inequality constraint, thereby changing the nature of the optimization problem. Fourthly, the methodology suggests the need for storage, i.e t, when the processing times suggest otherwise. Processes 1 and 3 commence at 0.5 h from the beginning of the time horizon, which is actually the completion time for process 2. This provides a direct reuse opportunity with no requirement for storage. The following sections of this paper demonstrate the applicability of the above-mentioned adaptations to completely batch operations using an hypothetical example of completely batch processes drawn from experience at an agrochemical facility. Initially, time is treated as the primary constraint and concentration as the secondary constraint. Subsequently, the priority of constraints is reversed so as to demonstrate the effect of the targeting procedure on the final design. Attention is also drawn to the fact that first sequence and cyclic-state targeting are essential in completely batch operations. The cyclic-state is assumed over an extended time horizon when more than one batches have to be produced in multi-stage operations. The first sequence corresponds to a single batch over a relatively shorter the time horizon of interest. In order to apply the presented method to a cyclic-state behaviour, the start and finish times would have to be provided for a much longer time horizon. 2. Graphical analysis for completely batch processes 2.1. Problem Statement Process 1 10 t Process 3 The problem addressed in this paper can be stated as follows. For each water using operation, given: 2.5 t 50 t Storage 37.5 t Process 2 40 t 12.5 t (i) the contaminant mass load, (ii) the fixed water requirement, (iii) the starting and finishing times to achieve the desired effect, e.g. mass transfer, degree of cleanliness of the vessel, etc. and (iv) maximum inlet and outlet concentrations, Water t Fig. 7. Resultant network for the Wang and Smith problem. determine the minimum amount of freshwater that can be achieved through the exploitation of reuse and recycle opportunities, as well as the concomitant water network. T. Majozi et al. / Journal of Environmental Management 78 (2006) It is worthy of note that freshwater minimization is concomitant with reduction in wastewater generation. It is also assumed that the considered processes are compatible, implying that product integrity is not compromised. This implies that the issue of product mixing is excluded Time taken as a primary constraint Taking time as primary constraint implies that at each stage during the course of the analysis, the concentration constraint is readily obeyed. To illustrate the application of this analysis to completely batch processes, an hypothetical example involving liquid liquid extraction (product washing) with water as the aqueous phase in the production of three agrochemicals A, B and C, was considered. These agrochemicals were produced in batch reactors. All three reactions formed sodium chloride (NaCl) as a byproduct, which was later, removed from the final product. The removal of this by-product was effected by the use of fresh water. It is worth mentioning that, although the aim of the washes was to remove NaCl, there were always traces of organics in water. In formulating the problem, however, it was assumed that the concentration of these organics was virtually negligible. In the case of A, the reaction took place in an organic solvent, which was highly immiscible with water, so that water was required solely for washing the salt. In the case of B and C, however, water was used as the reaction solvent, and a further quantity was used for washing the product. While investigating this secondary washing of B and C, it was found that the salt load removed from the product was essentially zero due to the fact that most of it had been removed with the reaction solvent water. However, it was considered that the washing step should not be discarded, as it constituted a quality control precaution in case of unforeseen process problems. The timing of the reaction and washing sequences was considered to be fixed by product requirements, which implies that there was no freedom to change the sequence to optimize the use of water Problem specification for graphical analysis The vessels operate in completely batch mode, implying that mass rather than flowrate of water is the relevant parameter. Setting the concentration limits presented some conceptual difficulties. The situation for product A was straightforward, as laboratory tests had shown that acceptable product quality could only be achieved if fresh water was used at the start of the wash. The reaction solvent water for B and C was also easily specified, since the only requirement was that the salt must not precipitate, which set the salt concentration at just less than 35% by mass at the end of the reaction. Precipitation of salt led to the erosion of glass lining in the reactors. The quantity of the wash water for B and C was based on experience. It was found that using water below Table 2 Problem specification for pinch analysis Process Time C in,max C out,max Water Salt load h kg salt/kg water Kg B reaction x x C reaction x x Total led to lower product yields. This could be due to inefficient mixing in the vessel. The quantity of water was then set at. The i
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!