(502g) Automated Targeting for Property Integration with Interception Placement

Traditionally, process industries have been focusing on conventional end-of-pipe treatment. However, over the past decades, the center of attention has shifted to more sustainable operations via effective use of resources. Amongst the few reasons that have resulted in this change include environmental sustainability, stringent emission legislation, as well as increasing of fresh resources and waste treatment costs.

Over the past decade, the application of mass integration techniques for material reuse/recycle has gained good attention among the academics as well as industrial practitioners. Recent reviews of mass integration can be found in literature (e.g. El-Halwagi, M.M., 1997, 1998, 2006; El-Halwagi and Springs, 1998; Dunn and El-Halwagi, 2003). However, mass integration techniques track individual chemical species without considering other properties or functionalities of the streams. In response to this limitation, the notion of property integration has been introduced by El-Halwagi and his co-workers, which takes into consideration the properties and functionalities of process streams.

Property integration is defined as a functionality-based, holistic approach to the allocation and manipulation of streams and processing units that include the tracking, adjustment, assignment, and matching functionalities throughout the process (El-Halwagi et al., 2004). In property integration, the material recovery problem is mapped to a cluster domain, which achieves the same objective as in mass integration with minimum resource usage and waste generation.

To further illustrate the concept of property integration, Shelley and El-Halwagi (2000) introduced a property-based cluster that was driven by tracking functionalities and properties of the streams instead of focusing on the chemical constituents. This technique obtains optimal recovery and allocation of volatile organic compounds in a complex hydrocarbon mixture. Later, Kazantzi and El-Halwagi (2005) introduced a pinch-based graphical targeting technique, which is generalised from the conventional material reuse/recycle pinch diagram (El-Halwagi et al., 2003). On the other hand, Foo et al. (2006) presented both graphical (property surplus diagram) and algebraic (property cascade analysis) techniques to locate minimum resource targets within a property integration framework. Most recently, Pau et al. (2007) extended the use of property cascade analysis (Foo et al., 2006) to determine rigorous targets for minimum fresh usage, waste discharge and interception targets of resource conservation network (RCN) with interception placement.

It is worth noting that property-based RCN synthesis has been explored in wide range of chemical processes. All the above mentioned works can basically be classified into insight-based and mathematical optimization techniques. To date, some works that utilised both insight-based and mathematical optimisation techniques have also been reported in the area of process integration. Among these works, the automated targeting approach presented by Ng et al. (2008) incorporated the targeting concept of insight-based technique into the mathematical optimisation model to locate the minimum flowrate/cost targets for a RCN. The proposed approach overcomes the limitations when both of the techniques are used independently.

In this work, the automated targeting technique (Ng et al, 2008) is extended for locating the minimum flowrate/cost targets for a property-based RCN. Two literature examples are solved to illustrate the proposed approach.

The automated targeting technique was developed by Ng et al. (2008) based on the concept of algebraic targeting technique of cascade analysis (Manan et al., 2004), with the removal of the dual-step procedure. It is worth noting that in all cascade analysis techniques, infeasible cascades with material flow balances are first generated to determine the largest material deficit, which is then added as fresh resource in the second step to remove all deficits and yield a feasible material cascade. Successful application is seen in water network, utility gas network and property-based network synthesis. Via the proposed automated targeting approach, the two-step targeting approach is readily replaced. In this work, the automated targeting is adapted for a property-based RCN.

The procedure for the automated targeting technique for a property-based RCN is next illustrated. A revised property interval diagram (Foo et al., 2006) is first constructed, where the property operators (Øk) of the material sinks and sources are arranged in an ascending order, from the lowest level k = 1 to the highest level k = n. In cases where the property operator levels for fresh resource(s) and zero property operator level (i.e. 0 MÙ-1 (Foo et al., 2006)) do not exist among the process sinks and sources, an additional property operator level is added. Besides, an arbitrarily highest property operator level is also added in last level of property interval diagram to allow the calculation of residue property load.

Next, material flowrate cascading across all property operator levels is to be performed. At each property operator level k (pk), the difference between the total available material sinks and sources (Sum FSRi - Sum FSKj) may be determined. Equation 1 shows the net material flowrate of each k-th level (Fk). As shown, Fk is the sum of the net material flowrate cascaded from the earlier property operator level, Fk-1 with the flowrate balance at property operator level k, (Sum FSRi - Sum FSKj )k, i.e.:

Fk = Fk-1 + ( Sum FSRi - Sum FSKj )k (Equation 1)

Note that the net material flowrate (Fk) can either take positive or negative value, with positive value indicates material that flows from the lower level into higher level or vice versa. This is in agreement with the cascade analysis technique.

Apart from material flowrate cascading, property load cascade is also essential to ensure a feasible RCN. Property load cascading from level k-1 to level k is performed as follows. Within each property operator interval, the property load is given by the product of the net material flowrate from level k and the difference between two adjacent property operator levels. Similar to the material flowrate cascade, residual property load of each concentration level k (Mk) is to be cascaded down to the next property operator level. Hence, property load balance at the k-th concentration level is determined by equation below:

Mk = Mk-1 + Fk(pk+1 - pk) (Equation 2)

where Mk-1 is the residual property load that is cascaded from concentration level k-1.

In order to ensure maximum allowable property load of sink in each level is fulfilled, and the property load is transferred from lower to higher level, the residual property load, Mk must take a positive value.

Mk >= 0 (Equation 3)

Therefore, Equation 3 is included as a constraint in the optimisation model. Note also that, a pinch point is observed where the residue property load is zero along the cascade.

In most cases, the fresh resource is of the highest quality, and corresponds to zero property operator. In cases where the fresh resource does not exist at the zero operator value, a new property operator level (pFR II) can be added. In order to determine the minimum total fresh resource flowrate, Sum(FFR + FFR II), the objective function for the mathematical optimisation model is formulated as: minimize Sum(FFR + FFR II) subject to the constraints in Equations 1 to 3.

It is worth noting that the above optimisation model is a linear programming (LP) problem, which can be solved easily to achieve global optimal solution if one exists. Two different case studies has been modeled and applied in LINGO software to prove the applicability of the methodology. These case studies have been used to demonstrate the flexibility of the automated targeting approach to analyse the interaction between the pre-treatment systems with the RCN.

REFERENCE:

1. El-Halwagi, M.M., 2006. Process Integration. Elsevier Inc., Amsterdam, Netherlands.

2. El-Halwagi, M.M., 1997. Pollution Prevention through Process Integration: Systematic Design Tools. Academic Press, San Diego.

3. El-Halwagi, M.M., Gabriel, F., Harell, D., 2003. Rigorous graphical targeting for resource conservation via material recycle/reuse networks. Ind. Eng. Chem. Res. 42, 4319-4328.

4. Foo, D.C.Y., Kazantzi, V., El-Halwagi, M.M., Manan, Z.A., 2006. Surplus diagram and cascade analysis technique for targeting property-based material reuse networks, Chemical Engineering Science 61(8), 2626-2642.

5. Kazantzi, V., El-Halwagi, M.M., 2005. Targeting material reuse via property integration. Chemical Engineering Progress 101 (8), 28-37.

6. Manan, Z. A., Tan, Y. L., Foo, D. C. Y., 2004. Targeting the minimum water flowrate using water cascade analysis technique. AIChE J. 50(12) 3169-3183.

7. Ng, D. K. S., Foo, D. C. Y., Tan R. R., 2008. Automated targeting technique for resource conservation networks. An international Conference on Water & Wastewater (Asia Water 2008).

8. Pau, C. H., Tan, Y. L., Ng, D. K. S. and Foo, D. C. Y. (2007). Property Integration Networks With Regeneration Placement, paper presented in Curtin University of Technology Sarawak Engineering Conference (CUTSE 2007), Sarawak.

9. Shelley, M.D., El-Halwagi, M.M., 2000. Componentless design of recovery and allocation systems: a functionality-based clustering approach. Computers and Chemical Engineering 24, 2081-2091.