Page 262 - Defrosting for Air Source Heat Pump

9
Defrosting control strategy

9.1 Introduction

For an ASHP unit during defrosting, it is easy to understand that its energy efficiency
is always affected by the defrosting initiation and termination control strategies [1].
The recent advancements in artificial intelligence and the Internet of Things also pro-
mote the developments of advanced control strategies for various equipment and pro-
cesses [2]. For an ASHP unit, two types of defrosting initiation control strategies are in
place, time-based and demand-based. For the first type, defrosting initiation is always
simply controlled by using a preset timer, at 60–90 min of frosting duration. However,
it is hard to give an exact and fixed frosting duration or defrosting starting point, due to
the complicated and changeable ASHP operating conditions. Therefore, the use of a
time-based initiation strategy always results in two typical mal-defrosting problems:
unnecessary defrosting when no or little frost is accumulated on the outdoor coil’s
surface, and no defrosting when it is actually needed. As shown by Wang et al., these
mal-defrosting phenomena adversely downgrade system frosting performance. For
example, after the ASHP system was operated for 5 days, the aforementioned second
type of mal-defrosting was found, with a frosted area of more than 60% on the mul-
ticircuit outdoor coil while no defrosting was initiated. Meanwhile, the system COP
was significantly reduced to only 2.3 while the ambient air temperature was at 7.9°C.
As analyzed with the testing results before and after the frosting operation, mal-
defrosting would decrease the COP and the heating capacity by 40.4% and 43.4%,
respectively [1, 3].
Demand-based defrosting initiation control strategy could start a defrosting oper-
ation only when sufficient frost is formed to adversely affect the operational perfor-
mance of ASHP units. Thus, it relies on accurately detecting the presence and growth
of frost by using direct and indirect frost accumulation-sensing technologies. These
technologies could be classified as: (1) measuring the frost thickness using a holo-
graphic interferometry technique; (2) measuring the frost surface temperature by
an infrared thermometer [4]; (3) detecting the refrigerant flow instability [5]; (4)
detecting the frost accumulation using a photocoupler or photo- or fiberoptic sensors;
(5) simulating the frost amount by applying neural networks [6]; and (6) calculating
the effective mass-flow fraction [7]. The frost accumulation on the surface of the out-
door coils was fully investigated with both experimental and modeling approaches.
For example, in Yao’s distributed mathematical frosting model, the frost thickness
°
could reach 0.5 mm at most with ambient air temperature at 1.5 C and relative humid-
ity at 85%. Based on the first principles, a frosting model was developed and validated
by Da Silva, and the effects of progressive frost clogging and the low conductivity of
the frost layer on the overall thermal resistance were assessed, suggesting that the for-
mer is the main cause of heating capacity reduction under frosting conditions [8].In
Ye’s study, the errors of numerical heat-transfer rate and frost mass with the
Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00009-6
© 2019 Elsevier Ltd. All rights reserved.