Optimizing the topologies of heating, ventilation, and air-conditioning water systems in supertall buildings: A pilot study

Abstract

The vertical topologies of HVAC water systems in supertall buildings are one major concern, which yet has been determined in experience-based manners. Therefore, the current article proposes a computational optimization procedure to search optimal design of system topologies, as well as component sizing and control settings. To reduce the computational efforts, the vertical topologies are simplified by considering only major components (e.g., air-handling units, heat exchangers, and chillers).Since both electricity consumption and capital cost are primary concerns for system design, the optimization problem has been solved using a two-objective solver, Nondominated Sorting Genetic Algorithm II. This optimization problem is then validated using a supertall building in Shanghai, and it is concluded that the generated solutions are superior in terms of both electricity consumption and capital cost. Finally, parametric analyses are conducted to provide insight into the uncontrollable variables (i.e., building location, building characteristics) with the hope that system design can be facilitated.

Nomenclature

Δp	=	pressure difference
k	=	pressure resistance coefficient
m	=	mass flow rate
C	=	capital cost
a	=	coefficient vector for polynomial curve
x	=	independent variables
J	=	objective function
P	=	power
I	=	user defined function
T	=	temperature
ε	=	tolerance
L	=	cooling/heating load
CT	=	cooling tower
AHU	=	air-handling unit
VFD	=	variable frequency driver
PHX	=	plate heat exchanger
SA	=	supply air
DH	=	dissatisfied hours
WP	=	working pressure
NV	=	the number of vertical zones
H	=	height or head
Cap	=	cooling/heating capacity

Subscripts
min	=	minimum value
SAT	=	supply air temperature
chi	=	chiller
des	=	design condition

Additional information

Funding

The current article was funded by the project in the National Science and Technology Pillar Program during the 13th Five-year Plan Period (2015BAL04B00).