S&W397-2LtSLL
CUNf-7709
53--
Preprint
fiom
the
26"
IEEE
Photovoltaic Specialists Conference,
Anaheim,
Sept.
1997
S
,4-
-9
7-2
y56c
AMP-HOUR COUNTING CHARGE CONTROL FOR
PHOTOVOLTAIC HYBRID POWER SYSTEMS
Thomas
D.
Hund
Sandia National Laboratories
PO Box 5800
Albuquerque, New Mexico 871 85-0753
ABSTRACT
An amp-hour counting battery charge control
algorithm has been defined and tested using the Digital
Solar Technologies MPR-9400 microprocessor based
photovoltaic hybrid charge controller. This work included
extensive laboratory and field testing of the charge
algorithm on vented lead-antimony and valve regulated
lead-acid batteries. The test results have shown that with
proper setup amp-hour counting charge control
is
more
effective than conventional voltage regulated sub-array
shedding in returning the lead-acid battery to a high state
of charge.
INTRODUCTION
Batteries in stand-alone and PV hybrid systems are
commonly subject to abusive conditions that are generally
due to,
1)
under charging in low resource periods, 2)
excessive charging in high resource periods, and 3)
inappropriate or ineffective charge control for the battery
technology. The individual
or
combined effects of resource
changes, poor charge control, and the daily load changes
can be potentially damaging to the battery. Previously
available PV charge controllers or charge control strategies
such as onloff PV array shedding will generally provide the
battery with sufficient charging current to complete the bulk
charge phase which will return the battery to 80
to
95%
state of charge (SOC) [1,2]. After the bulk charge phase,
the taper
or
absorption charge phase is very important in
preventing stratification, hard sulfation, and premature
capacity
loss.
If a regulation voltage of 2.35 to 2.40 vpc is
used for charging vented lead-antimony batteries, then a
10 to 24-hr regulated voltage finish charge period is usually
required. The available time for battery finish charging in
PV systems is generally much less than 10-hrs. The short
time at regulation voltage results in an incomplete finish
charge phase which consistently leaves the battery in an
under charged condition. If the regulation voltage for
vented lead-antimony batteries is raised to 2.45
to
2.50 vpc
and the reconnect voltage is raised to 2.28 to 2.30 vpc,
then the finish charge time period can be significantly
reduced. With the higher regulation voltage, the battery
may then be subject to excessive charge in high resource
periods or low load periods. The primary effect of the
higher regulation voltage without amp-hour counting
charge control on vented batteries is the dramatically
Bruce Thompson
Biri Systems
30 Sodom Rd.
Ithaca, NY 14850
increased watering requirements and increased erosion of
active plate material. For VRLA batteries the regulation
voltage should be set to the manufacturers recommended
value for a cycling application. VRLA battery cycling
regulation voltage is usually 2.35 or 2.40 vpc. A reconnect
voltage of 2.28 to 2.30 vpc
is
also very important to
complete the finish charge period. Very little can be done
to accelerate the finish charge period for VRIA batteries
because their charge acceptance is limited by the oxygen
recombination cycle
[3].
Higher regulation voltages will
only accelerate dry-out. That is why it is
so
very important
to use constant voltage charging for the finish charge
phase or simulate it with a well designed PV array
shedding strategy.
Based on energy calculations from the "RAPS Design
Manual", published by the University of Cape Town South
Africa and the author's own calculations, battery energy
costs for PV hybrid systems are estimated to be about
$0.35 to $l.OO/kWh over the life of the system [4,5]. As
indicated above, any degradation in battery cycle-life can
result in a significant system cost increase. The potential
cost benefit to stand-alone and PV hybrid systems is
substantial if batteries meet their rated cycle-life.
Amp-hour (Ah) counting charge control for PV hybrid
battery charging systems is new to this application, but in
the Battery Technical Manual from Battery Council
International the cycle-life test procedure for deep cycle
marine/RV batteries does use Ah counting as a means to
ensure the battery
is
at a high
SOC
[6].
In
this test
procedure the maximum and minimum values for percent
overcharge per cycle for vented and VRIA batteries are
specified.
This along with regulation voltage insures
complete battery recharge.
Maximizing battery cycle-life requires using the
manufacturers recommended regulation voltages,
appropriate system design, effective charge control, and
reaching the recommended overcharge
in
Ah when the
battery receives a full recharge. The excess Ah are a way
to compensate for battery efficiency losses. For most
vented lead-antimony batteries the recommended Ah
overcharge is between 120 and 130% of the discharged
Ah. VRIA batteries, which require more time at regulation
voltage, are much more efficient and only require between
105 and 115% of the discharged Ah. This paper will
evaluate an Ah counting charge control algorithm for PV
hybrid systems using a microprocessor based charge
controller.
This work is supported
by
the Photovoltaic Energy Technology Division,
US
Department
of
Energy.
Sandia
is
a
multiprogram
laboratory
operated
by
Sandia
Corporation,
a
Lockheed
Martin
Company. for
the
United
States
Depanmrnt
of
Energy
under
contract
DE-ACO-1-WAt85000.
DISCLAIMER
This report was prepared
as
an account of work sponsored by an agency of the United
States
Government Neither the United States Government nor any agency thereof, nor
any of their employees, make any warranty,
express
or
implied,
or
assumes
any
legal
liabili-
ty
or
responsibility
for
the accuracy, completeness,
or
usefulness of any information,
appa-
ratus,
.product,
or
process
disdased,
or
represents
that
its
use
would not infringe privately
owned
rights.
Reference herein
to
any
specific
commercial product, process,
or
service
by
trade
name,
trademark, manufacturer,
or
otherwise does not
necessarily
constitute
or
imply its endorsement, recommendation,
or
favoring by the United States Government
or
any agency thereof. The views and opinions of au!hors expressed herein do not necessar-
ily state
or
reflect those of the United States Government
or
any agency thereof.
AH COUNTING CHARGE CONTROL
In a cooperative effort with Digital Solar Technologies
a microprocessor based Ah counting charge control
algorithm was defined and tested using the MPR-9400 PV
hybrid charge controller. The controller under test uses
staged PV sub-array switching
to
achieve a taper charge
for the battery finish charge period. This charge control
method can be a very effective charge strategy if properly
setup.
It
is typically used in medium to large PV hybrid
systems.
The new Ah counting charge control algorithm
calculates battery Ah for each complete cycle. A new
cycle is started when the battery reaches the
predetermined overcharge in Ah. To implement the new
Ah counting charge algorithm required four new input
variables. These variables are:
1)
BATAHINIT
-
Estimated battery capacity
in
Ah (Input
by user),
2)
AHVRESET
-
Battery voltage when battery charging
or high voltage disconnects (HVD’s) are reactivated
(Input by user),
3)
%ADD
-
Deficit or excess battery Ah at initial battery
regulation voltage (Input by user
-
+or-25%),
4)
%OVER
-
Maximum overcharge above the daily Ah
DOD (Input by user
-
0
to
99%).
The Ah counting charge control algorithm opens the
high voltage disconnect relays (HVD
1
and
2)
when the
specified %OVER plus %ADD Ah are charged into the
battery. The %OVER and %ADD Ah begin counting when
the first HVD is reached. Before the specified overcharge
is reached, HVD
1
and
2
will operate as indicated by the
preset disconnect and reconnect voltages. The %OVER
value is defined in equation
1.
((Ah chr
-
Ah dischr) /Ah dischr) x
100
(1
1
Ah calculations for battery
SOC
and %ADD are
based on the battery BATAHINIT capacity input by the
user. The displayed battery capacity in Ah and battery
%SOC
are reset
to
the BATAHINIT and
100%
values
respectively when the required overcharge is reached, or
at
6
PM after an equalization charge. If a load is turned on
after charging is terminated, then the PV array will be
reconnected when the battery reaches the preset
AHVRESET voltage.
SETUP AND INITIALIZATION
The MPR-9400 initialization for Ah counting charge
control requires the user
to
identify the additional variables
from the previous section. The HVD
1
and 2 regulation
voltage
is
generally available from the battery
manufacturer and the reconnect voltage is usually set
based on the sub-array shedding sequence and system
design. The %OVER, and BATAHINIT values are
available from the battery manufacturer. The %ADD
parameter is very system dependent and can only be
accurately identified by:
1)
equalizing or boost charging the batteries,
2) resetting BATAHINIT
3)
4)
5)
running a standard cycle to regulation voltage,
recording the battery Ah when regulation voltage is
reached, and
calculating %ADD by equation 2.
((BATAHINIT
-
Bat Ah at Vr)
/
BATAHINIT) x
100
(2)
In most cases the %ADD variable will be less than
plus or minus
7%.
Test results have shown that as vented
lead-antimony batteries age they will require adjusting the
%ADD parameter from a positive number
to
less than
-
7%.
This is because of the increased gassing current
resulting from antimony poisoning
m.
VRLA batteries
typically will not experience significant changes in the
%ADD value over their life because they generally have a
stable end of charge current with very low gassing current.
PV SYSTEM SIMULATOR TEST RESULTS
The measured percent overcharge from three
different stand-alone PV system simulations and a PV
hybrid simulation has been obtained using an automated
PV hybrid system tester from Team Specialty Products in
Albuquerque. This test-bed is capable of simulating two
PV arrays or one PV array and one engine generator with
two
loads. All system parameters are fully programmable
and automated.
Stand-Alone
PV
System Simulation
The stand-alone PV system test was conducted
to
evaluate Ah counting charge control using an array Ah to
load Ah
(C:L)
ratio of
1.5,
1.75,
and 2.0. The simulation
was configured to duplicate a typical stand-alone PV
system using two 7-amp PV sub-arrays for battery
charging. The battery used in the system simulation was a
GNB
12-5OOOX
400 Ah VRLA battery. This battery is very
similar
to
the GNB Absolyte
IIP
which
is
commonly used
in
larger stand-alone and PV hybrid systems. A 1.5-amp
continuous load and an adjustable nighttime load was used
to establish the three different
C:L
ratios. The new Ah
counting charge control parameters for the MPR-9400
were input based on battery manufacturers specifications
and the %ADD calculation from the above initialization
cycle. See Table
1
for a complete list of system
parameters.
In Fig.
1
are the test results, which show that the Ah
counting charge control algorithm limits battery overcharge
to within the
105
to
110%
specified by GNB. The voltage
regulated charge period is also maintained between 1.9
and 2.3-hrs. Using the same regulation voltages and
system configuration without Ah counting charge control
resulted in a battery overcharge range between 114 and
128%.
The voltage regulated charge period
also
varied
between 3.9 and 5.0-hrs depending on the
C:L
ratio.
Without Ah counting charge control the only way to reduce
overcharge is to lower regulation voltage or provide a two
stage voltage regulated charge control. The lower
regulation voltage and two-stage charge control will
decrease overcharge, but it
will
also increase the time
required to charge the battery. Voltage regulated charge
2
control alone makes it very difficult to obtain optimum
Sub-Array 1
Sub-Array 2
Load 1
Load 2
Battery
Size
Battery DOD
C:L Ratio
MPR-9400
HVD-1
HVD-2
Temp. Comp.
BATAHINIT
AHVRESET
%ADD
%OVER
recharge due
to
weather and system design.
7-amp peak
7-amp peak
Continuous 1.5-amps
19.3, 15.3,
&
12.3-amps
400
Ah
13.5,
12.0,
&
10.5%
1.5,
1.75,
&
2.0
2.36 vpc Vr
2.35 vpc Vr
(-5mvlClcell)
400
Ah
2.08 vpc
1.9
7
2.30 vpc Vrr
2.29 vpc Vrr
Isvstern
I
Value
time of 10.7-hrs returning 152 Ah back into the battery.
The end of charge current, which can be used
to
terminate
charging, was measured at 3.9-amps.
425.0
400.0
375.0
S
U
350.0
2
325.0
rn
300.0
275.0
250.0
Fig.
2.
PV hybrid simulation using Ah counting charge
control.
ler setup
FIELD TEST RESULTS
105
.
.
.
- -
-.
100
7
I
1.25
I
.5 1.75
2
2.25
Amy
Ah
To
Load
Ah
Ratio (CL)
Fig.
I.
VRLA battery
%
overcharge with and without Ah
counting charge control.
PV
Hybrid Simulation
The PV hybrid simulation was designed with one 7-
amp PV array and an engine generator set to charge the
battery at 20-amps once the 12.1 volt engine start voltage
was reached. The Charge Controller was programmed to
terminate all charging when the required 107% overcharge
was reached, The system C:L ratio with one PV array was
0.62 and the %ADD input variable was still +1.9. This
system design resulted in the battery discharging
to
12.1
volts during the fourth cycle. The engine generator then
returned the battery
to
full charge after the fourth cycle.
In Fig. 2 are the battery Ah and voltage plotted over a
ten cycle period. The measured overcharge values after
the engine generator battery recharge were 110 and
I1
1%. The results indicate good reproducibility and they
are close to the desired value of 107%. This test is also
useful in identifying the effects of different PV array C:L
ratios and engine generator control strategies. In this case
the generator ran for a total of 3.5-hrs in bulk charge mode
and 7.2-hrs in finish or taper charge mode for a total run
In addition to the laboratory testing, the Ah counting
charge control algorithm was used on two stand-alone PV
systems with a 700 Ah Trojan L-16 vented lead-antimony
battery and a 400 Ah GNB 12-5OOOX VRLA AGM battery.
These system tests represent each lead-acid battery
technology and demonstrated that the Ah counting charge
control algorithm will function well with each battery type
even though they are significantly different in their
recharge characteristics. The most significant difference
between the
two
battery types is the battery Ah at initial
regulation voltage. This charging difference is
compensated for by the %ADD parameter in the MPR-
9400.
Ah
Charge Control Using Trojan L-16 Batteries
The stand-alone PV system using Trojan L-16
batteries is configured with three PV sub-arrays charging
at a peak current of
16,
10, and 4-amps. The two higher
current sub-arrays are switched with the MPR-9400 charge
controller and the 4-amp array is permanently connected to
the battery bank. The load is variable and is operational
day or night. Daily loads average 70 Ahfday but can range
between 10 and 200 Ahfday. The voltage setpoints on the
MPR-9400 are set for staged sub-array switching at:
HVD-1
=
2.45 vpc Vr and 2.29 vpc Vrr,
HVD-2
=
2.43 vpc Vr, and 2.27 vpc Vrr.
The Ah counting program parameters are set to:
BATAHINIT
=
300
Ah
AHVRESET
=
2.06 vpc
%OVER
=
30
%ADD
=
-1
1.7
%OVER (0.30~70)
+
%ADD
=
-14.1 Ah
In Fig. 3 are the Ah overcharged and the respective
percentage overcharge for a 15-day period. The results
3