Hydrocarbon Processing - March 2022 - 83

Water Management
H. BHATIA, C. GARCIA and K. NANGIA,
Fluor Enterprises Inc., Aliso Viejo, California
Optimize a cascaded cooling water network
Waste heat in process units is removed via heat exchange with
a recirculating cooling water supply and return stream (FIG. 1).
The heat picked up by the cooling water exiting the network of
heat exchangers is removed via direct contact with ambient air
in a direct contact cooling tower.
While numerous strategies and publications exist related
to optimizing the cooling tower operation, this article focuses
solely on the optimization of cooling water networks within a
refinery process unit.
In the past, process engineers have given a lower priority to
optimizing cooling water networks. Many process units either
used more cooling water and energy than necessary or supplied
a surfeit of cooling water to a few users at the expense of failing
to deliver sufficient cooling water to many others.
The optimization of a cooling water network has now become
a major design consideration for improving the operability
and profitability for the end user. The objective of this
analysis is to reduce the operational expenditure (OPEX) and,
therefore, increase profitability. In many instances, this type of
optimization has also resulted in reducing capital expenditure.
An analysis is presented here of a cooling water network for
a refinery process unit initially considered for implementation,
and the subsequent network optimized using commercially
available hydraulic analysis software. While the case study analyses
presented are for a specific process unit in a grassroots refinery
installation, the concepts are fundamental in nature and
may be broadly applied to any cooling water network.
Cooling water networks. In a typical network, all services are
supplied with cooling water at the same temperature (FIG. 2).
FIG. 2. Typical cooling water network.
While the supply pressure at each user may vary depending on
the elevation of the exchanger above grade level, the pressure
in the cooling water return header is usually the same (or very
similar) for all users.
Such an arrangement demands a substantial amount of water
circulation. Moreover, due to the maldistribution caused by differences
in elevations above grade and hydraulic resistances in
piping and/or heat exchangers, a prudent process design will
often include a sizable design margin (20%-25%) over and
above the initial cooling water utility balance, as determined
from process datasheets.
The above approach is a classic example of how cooling water
networks were designed in the past. However judicious this
may appear, it fails to take advantage of the fact that some services
may accept a higher supply temperature with negligible or
minimal thermal impact. In such a scenario, a secondary cooling
water header (FIG. 3) is supplied with cooling water " cascaded "
from some services supplied by the " primary " cooling water
header-this can be considered a means of reducing the cooling
water recirculation rate.
Though fundamentally simple in concept, the cascaded cooling
water network requires a more elaborate hydraulic analysis
due to the combination of a parallel and series-parallel network
FIG. 1. Recirculating cooling water system.
FIG. 3. Cascaded cooling water network.
Hydrocarbon Processing | MARCH 2022 83
Hydrocarbon Processing - March 2022

Table of Contents for the Digital Edition of Hydrocarbon Processing - March 2022
