Fig。 2。 Lookup table data for the variation of air-conditioner part load sensible cool- ing ratio as a function of ambient temperature, building return air dry bulb and building wet-bulb temperature。

t s

Values of Qt/Q ∗  and Pac /Pac ∗  as a function of ambient    temper-

ature and building return air wet-bulb temperature are shown in Fig。 1。 Data is taken from Klein et al。 [42]。 The value of Qs/Q ∗ varied as a function of ambient temperature and both the building  return

air dry and wet bulb temperatures as shown in Fig。  2。

The air-conditioning requirement was controlled according to the varying building temperature Tb  with a hysteresis such that    if

If the supply air humidity ratio calculated using Eq。 (2) was greater than the saturation humidity ratio at the estimated supply air temperature, then the supply air was assumed to be saturated and the value of Ts  was adjusted so that Qt  = m˙   (hb − hs)。

The following performance values at the rated conditioned were

used; Q ∗ = 3kW , Q ∗ = 2。3kW and Pac ∗ = 0。6kW giving EER∗ = 5。

t s

Tb  > 25◦ C  then conditioning was required and if Tb  < 22◦ C  then

conditioning was not required。 If conditioning was required and sufficient power was available (see Section 2。2), then given Qt and Qs at the specific ambient and building condition, the conditioned supply air temperature Tsand humidity ratio HRs were estimated using the expressions;

  Qs

2。2。 PV-Battery system

The power output of the crystalline silicon photovoltaic array Ppvwas modelled using the approach described in Huld and Amillo [24] in which the power output is calculated from the incident total radiation on the collector surface taking into account the  influence

Ts  = Tb  − m˙  c

(1)

of module temperature on performance。 The module   temperature

was evaluated as a function of the ambient temperature and   wind

HRs  = HRb  − (Qt  − Qs) /m˙  hfg (2)

speed over the array。

Charging and discharging of the battery system was modelled

where the supply air flow rate m˙

= 0。275kg/s

using an adapted version of the approach described by Tant et al。

M。J。 Goldsworthy / Energy and Buildings 135 (2017) 176–186 179

[25]。 If Ppv > Pac then the battery charging power was calculated as the minimum of the available power and the maximum charging

power according toPc = min 。μc Enom, Ppv − Pac 。 where μc is the bat-

tery charging power to energy ratio and Enom is the nominal battery energy storage capacity。 If Ppv  < Pac  then the battery    discharging

power was calculated as Pd  = min 。μdEnom, Pac  − Ppv。 where μd   is

the battery discharging power to energy ratio。 If Ppv + Pd < Pac , then the air-conditioner was assumed to be off and the battery was

charged according to the charging power equation  given above。 The energy stored in the battery Ebat was updated according to the equation;

dEbat

dt   = yc Pc − 1/ydPd (3)

where yc and yd are the charge and discharge efficiencies respec- tively。

If, at any time the energy in the battery exceeded 80% state of charge, no further energy was added to the battery。 Similarly, if the battery state of charged dropped below 5%, then no further energy was drawn from the battery。 At the start of the calculation, the battery was assumed to be at 40% of its nominal energy capacity。