(699h) Proactive Optimization-Based Reconfiguration of Heat-Exchanger Super Networks

Large scale chemical or petrochemical processes are highly energy intensive and usually require energy recovery systems to guarantee high efficiency and low cost. A heat-exchanger network (HEN) aims at utilizing internal flow streams to satisfy outlet temperature targets so that operating costs can be minimized. A traditional HEN is expected to be operated under the condition of maximum heat integration and minimum utility consumption [1]. An emerging trend in energy management is to integrate renewable generation, such as wind and solar resources and to supply heat and power not only to residential users but also to process industries. As a result, the electricity can be “zero-cost” as well as “zero emission”, making renewable energy a promising power alternative, although the discrete generation and inherent uncertainty in renewable resources cannot be neglected. Furthermore, there is a trade-off between energy recovery using a HEN and adding external (auxiliary) energy to the process, the optimal decision of which depends on the corresponding costs.

Most of the control and operation research regarding HENs assumes that the synthesis is fixed, which means all necessary network exchangers, utility units, and the network structure are well defined [2]. In this work, a supervisor level is proposed that plays a dominant role to decide the operating states of the local HEN according to varying electricity pricing and available heat-exchangers. This supervisor explores the space of possible optimal HENs and finds optimal structures within a superstructure that satisfies operational and economic targets and constraints. This superstructure can be proactively optimized and implemented through a receding horizon approach.

The cost function for utilizing network heat-exchangers and utilities powered by electricity from either renewable resources or the grid is expressed as dollars per kWh. The more energy transferred within the HEN, the lower the amortized cost is with respect to per unit of energy due to the scaling effects for the equipment. On the other hand, the cost function, when multi-step tariff pricing is considered, would increase when the renewable generation is scarce and the utility has to be powered by electricity from the grid. An economic objective function considering both capital and operating costs is thus defined and optimized to decide which HEN structure is optimal at any given time period and also to determine the set-points for local controllers to follow. This hierarchical structure takes advantage of the disparate time scales associated with varying pricing dynamics (low frequency) and local set-point tracking (high frequency).

Since heat-exchanger performance can be complex, the prediction of their operation from first principles can be challenging.  The feasible synthesis of HEN are presented using advanced process simulations tools (such as Aspen Plus^®), considering both the steady-state and dynamic models to incorporate controllers.

This paper will present our preliminary results in defining proactive heat-exchanger super structures where transitions between HEN structures are instantaneous. We envision incorporating dynamic models not only for the local control scheme but also in the supervisory level to account for the structural transitions that incur operational and possibly capital costs. 

[1] González A H, Odloak D, Marchetti J L. Predictive control applied to heat-exchanger networks. Chemical Engineering and Processing: Process Intensification, 2006, 45(8): 661-671.

[2] Aguilera N, Marchetti J L. Optimizing and controlling the operation of heat-exchanger networks. AIChE  Journal, 1998, 44(5): 1090-1104.

[3] Gorji-Bandpy M, Yahyazadeh-Jelodar H, Khalili M. Optimization of heat exchanger network. Applied Thermal Engineering, 2011, 31(5): 779-784.

[4] Zamora J M, Grossmann I E. A global MINLP optimization algorithm for the synthesis of heat exchanger networks with no stream splits. Computers & Chemical Engineering, 1998, 22(3): 367-384.