Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02643684 2008-10-28
Patent Application of
Ahmed Abdelsamie and Steve Carkner
For
TITLE: A MULTI-BATTERY SYSTEM FOR HIGH VOLTAGE
APPLICATIONS WITH PROPORTION POWER SHARING
CROSS-REFERENCE TO RELATED APPLICATIONS
None.
BACKGROUND - FIELD OF THE INVENTION
This invention relates to a method or apparatus for controlled or regulated
charging,
discharging, or combined charging and discharging of one or more voltaic
cells, batteries,
or capacitors. Specifically, this invention relates to a multi-battery system
for high-
voltage applications with proportional power sharing.
BACKGROUND - DESCRIPTION OF THE PRIOR ART
A number of operationally critical applications have power requirements that
require high
voltages and currents and require a high degree of availability and
reliability. One way of
ensuring high reliability is to have a high degree of redundancy in as many
components
of the system as possible. In battery operated systems the most critical
component is
often the batteries and in high voltage applications a large number of
batteries may be
required to produce the necessary power output. For example, in an
uninterruptible
power supply (UPS) the system relies on multiple batteries feeding a voltage
converter
1
CA 02643684 2008-10-28
system that increases the system voltage. The UPS relies upon switched mode
power
supplies formed from inductors, capacitors and other elements in order to
convert an
incoming DC voltage to an output DC or AC voltage. In cases where redundancy
is
required, multiple batteries and multiple voltage converters are used to
provide power in
the event that a single system element fails. The requirement for redundancy
can result in
very large and unwieldy systems due to the duplication and even triplication
of system
components. Where multiple battery banks are relied upon to supply an input
voltage the
system may not share the charge equally between battery banks. This results in
one
battery bank or one voltage converter supplying the majority of the standby
power.
Ideally, the system should be able to determine which battery bank has the
highest state
of charge in order to supply the most reliable input voltage. The inability to
do so may
result in the UPS continually relying upon a depleted battery bank for input
voltage.
Overtime, the depleted batteries will fail to charge and the UPS becomes less
reliable.
To partially overcome this design deficiency power systems relying upon
multiple
batteries power share between them. This results in equal power consumption
from each
battery. However such systems can still fail to take into account the state of
charge of the
batteries. This will result in a depleted battery continually being called
upon to share a
load it is not capable of producing. In the end, the battery will fail and the
system
reliability will be compromised.
Another situation can occur in very large power systems where perhaps hundreds
of
batteries are used. Individual batteries or groups of batteries can be "hot-
swapped". This
2
CA 02643684 2008-10-28
results in the battery bank containing both older and new batteries having a
variety of
charge states and reliabilities. Overall, system reliability is reduced. In
this situation, a
system that is able to detect the state of charge of a battery or battery bank
and rely upon
the most fully charged batteries to supply input voltage would ensure longer
battery life
and a greater UPS reliability.
Therefore, in light of the deficiencies in UPS systems noted above, there is a
need for a
multi-battery system that provides redundant input voltage without the expense
and bulk
of duplicated or triplicate critical components.
SUMMARY
In accordance with the present invention there is provided a multi-battery
system for
high-voltage applications having a proportional power sharing scheme. In one
embodiment of the present invention there is provided at least two batteries
coupled to a
respective at least two energy storage means. The energy storage means are
controlled by
a master monitoring circuit connected at the point of load. Each energy
storage means
may be additionally controlled by a monitoring circuit on its respective
battery. The
energy storage means will store more power when connected to a battery having
a higher
output voltage and hence a higher state of charge. A battery with a lower
state of charge
will experience a more severe voltage drop under heavier loads. This is
detected by the
monitoring circuit and the system will be manipulated to favor the battery
having the
highest state of charge. It can be appreciated that the system of the present
invention may
comprise an infinite number of batteries combined with an infinite number of
storage
3
CA 02643684 2008-10-28
elements. Where the system comprises N batteries and where the total demanded
power
output of the system is P, then the total power that each energy storage means
must carry
would be equal to P/N. Each energy storage means may be designed to handle a
larger
fraction of the total required power. For example, if each energy storage
means were
designed such that it handled 2P/N, then only half of the energy storage means
would
need to function in order to supply the full power requirements of the system.
In this way
system reliability is increased without having to increase component
redundancy.
OBJECTS AND ADVANTAGES
Accordingly, besides the objects and advantages of multi-battery system
described
herein, several objects and advantages of the invention are:
a. There is no need to install expensive redundant systems to create a system
that
will provide a high degree of reliability in the event of system component
failures.
b. The system automatically allows multiple battery systems to share the power
load.
c. Higher states of charge batteries carry a higher proportion of the demanded
load than lower state of charge batteries.
4
CA 02643684 2008-10-28
DRAWINGS AND FIGURES
Figure 1 is a schematic of an example of a known system for generating high
voltages
from a battery supply.
Figure 2 is a schematic of one known method to increase system reliability.
Figure 3 is a schematic of another known method for increasing system
reliability.
Figure 4 is a schematic of yet another known method for increasing system
reliability.
Figure 5 is a schematic of one embodiment of the invention.
Figure 6 is a schematic of another embodiment of the invention.
DESCRIPTION
Known systems and their deficiencies
Referring to Figure 1, there is shown, as an example, a schematic drawing of a
known
system 98 for generating high voltages from a battery. The system comprises a
battery
100 comprising a plurality of battery cells. The voltage generated by the
battery will pass
through protection means 101 and be applied to the load 102 across the
terminals 104 and
106. If the load requires a voltage of Y volts and each battery cell has a
typical voltage of
X volts then the total number of cells required in this system can be
calculated by X/Y.
Since the cells are series connected, the failure of any one cell will result
in failure of the
battery. As well, the failure of the protection means will also fail the
system. Hence, this
system is very unreliable.
As the battery 100 discharges the output voltage will vary over time and this
will be
detrimental to the load 102. Reliability of the system can be improved by
regularly
5
CA 02643684 2008-10-28
changing the used battery 100 with a fresh battery. Alternatively a second
battery can be
added to the system to increase redundancy. However, both alternatives are not
cost or
labor efficient. In another embodiment of this known system 98, a circuit
could be added
to each cell of the battery to detect cell failure and subsequently isolate
the cell from the
battery. However, the addition of circuitry to each cell of a large battery
would result in a
costly system and high resistance loads on the battery from the circuitry
itself.
Referring to Figure 2, there is shown another example of a battery system 198
for
generating high voltages having greater reliability than the system of Figure
1. In this
system there are multiple smaller battery packs 200 to supply the load. This
triple
redundancy increases the reliability of the voltage source. The each battery
circuit
comprises a battery pack 200 in series with protection means 201 and voltage
blocking
element 202. In the event that a battery pack fails, the blocking element will
isolate the
failed battery circuit from the other two surviving circuits. The current from
each battery
circuit is then combined before entering a voltage converter 203 which takes
the low
voltage of the battery packs and increases it to the required system output
voltage 204
required by the load across the terminals 206 and 208. The advantage of this
system is in
the redundancy of the battery circuits. Any one battery circuit (or two
circuits) can fail
and the system continues to function on the surviving circuit(s). A further
advantage of
this system is the ability to regulate the voltage output 204 regardless of
the state of
charge of individual batteries. A disadvantage of this system is the
requirement for all of
the system power to pass through a non-redundant voltage converter 203.
Failure of the
voltage converter would fail the system. As well, the system will require
heavy cable
6
CA 02643684 2008-10-28
connections between the voltage converter and the batteries to carry the
current
demanded by the load. Another major disadvantage of this system is its
inability to select
the battery circuits having the greatest state of charge. One battery circuit
may end up
supplying more of the load than the other two circuits depending on the health
of the
battery cells, cell voltage, cell impedance and impedance of the series
components such
as protection means 201 and blocking element 202. Another disadvantage of this
system
is the threshold of the blocking element 202 which may be only a fraction of a
volt. In
that case, a battery that has a terminal voltage that fraction of a volt
higher than all other
batteries in the system will end up providing all of the power. The blocking
elements will
block current from the two lower voltage circuits. This situation is
especially dangerous
in systems where a freshly charged battery may be hot-swapped into an active
system.
Referring to Figure 3, there is illustrated a schematic of another known
method to
increase reliability of UPS systems. In the system illustrated 302, each
battery 200 in the
multiple battery system is serially connected to an independent voltage
converter 301.
The advantage of this system is that the failure of any battery or any voltage
converter will not render the system inoperable. The disadvantage of this
system lies in
the tendency for the voltage converters 301 to fail to share the system load.
If one voltage
converter senses the output 204 as being out of tolerance, it may attempt to
supply all the
power required to bring the output back into regulation. This problem is
exacerbated at
high loads as the distance from each battery to the load will result in larger
measurement
errors and therefore the batteries closest to the load are more likely to
supply more of the
power to the load. This makes it very difficult to balance power distribution
across the
7
CA 02643684 2008-10-28
batteries 200. This problem gets progressively more difficult as more
batteries are added
to the system. The state of charge of the individual batteries is also ignored
in this setup.
The amount of power delivered by any individual battery is based solely on the
accuracy
of the control circuitry involved in response to the load.
Referring now to Figure 4 there is illustrated a circuit 400 that shows an
example of a
slightly improved approach to improving reliability of multi-battery systems.
In this
system there are multiple batteries 200 and multiple voltage converters 401
serially
connected to each of the batteries. This circuit includes an analog feedback
system 403
that permits each of the voltage converters to sense the overall output
voltage 204. The
analogue feedback system ensures that each of the independent voltage
converters 401
will sense the identical output voltage 204. This improves accuracy of voltage
regulation
but using an analogue signal will have inherent inaccuracies due to how the
converters
interpret the signal. This may result in a single voltage battery circuit
delivering more or
less power to the load than the other two battery circuits. This circuit also
suffers from
inaccuracy because when many batteries are used in a single system the
probability of
error increases and the distribution of an accurate and noise-free analog
signal becomes a
challenge. This circuit also ignores the state of charge of the batteries and
instead seeks
only to balance the amount of power equally among all batteries.
8
CA 02643684 2008-10-28
Description of One Embodiment of the Present Invention
Referring now to Figure 5, there is shown one embodiment of the present
invention that
overcomes the deficiencies discussed above. The invention comprise a circuit
498 that
comprises a plurality of battery circuits each comprising a battery 500 that
is connected to
and monitored by protection means 501. Generally the circuit will have at
least two
batteries and each battery will have a battery positive 510 and a battery
negative 512
terminal. This combination of batteries will provide the system with the
desired level of
voltage output and reliability. Protection means 501 is connected across the
battery
positive and negative terminals. The output of the protection means is through
protection
means positive 514 and negative 516 terminals. The protection means output is
feed into
energy storage means 502 connected to the positive terminal of protection
means. The
energy storage means stores energy when it receives a true digital control
signal from a
logic circuit means 503. The logic circuit means is a gate that outputs a true
digital
control signal when the protection signal 505 from protection means 501 and
the digital
enabling signa1506 from voltage sensing element 504 are both true. The
enabling signal
is generated by a voltage sensing element 504 connected across the circuit
positive and
negative terminals. The voltage sensing element outputs a pulse-width
modulated or
frequency modulated digital enabling signal that varies with the output 204.
If the output
is too low, the enabling signal will be true for a larger proportion of time
which will
cause more energy to be stored in the energy storage means 502. The energy
storage
means 502 will discharge its stored energy when the digital input to it is
false. The energy
storage means will typically be composed of an inductor and at least one
transistor.
9
CA 02643684 2008-10-28
Other elements can be added. In another embodiment of the invention the energy
can be
stored in capacitors. Energy storage tends to be proportional with voltage. In
the case of
an inductor when a battery is connected to an inductor, current will ramp up
through the
inductor at a rate that is proportional to the inductance and the circuit
resistance. The
current is therefore essentially independent of voltage. The actual power
being stored is
therefore proportional to the voltage of the battery since power is equal to
voltage
multiplied by current. Similarly, if the energy storage means is a capacitor,
the energy
stored is equal to one-half capacitance multiplied by the square of the
terminal voltage.
Therefore, in the case of a capacitor storage element, a battery with a higher
voltage will
deliver considerably more power to the load due to the squared dependence on
voltage.
In this way, the battery 500 with the highest potential, which generally
relates to state of
charge, will also deliver the most energy to the load 204. This system
therefore
automatically, and without any specific analog or digital control, has a
tendency to cause
the battery with the highest capacity to also supply the most power to the
load.
Referring to Figure 6, power delivery to the output 204 can achieve lower
noise by
implementing separate controls to each battery 500 or by grouping the controls
such that
one energy storage means 502 would be discharging into the load while another
energy
storage means on another battery is charging. If the operating frequency of
the enabling
signal from the voltage sensing element 504 is F, then a system with two
separate control
groups would have an output noise frequency of 2F. Higher output frequencies
result in
lower total noise as filtering elements become more efficient at eliminating
noise as
frequency increases. It is possible to have as many control signals as there
are batteries.
CA 02643684 2008-10-28
The system would time shift each control signal the same amount. For example,
if there
were two enabling signals 506 from two voltage sensors 504 and the system was
operating at an update frequency of 1kHz in a pulse width modulating mode,
then the
period of the signal is 1/frequency or about 1 milli-second. The two enabling
signals
would therefore have similar pulse width and identical frequency, but would be
time-
shifted by 0.5 milli-seconds. If four enabling signals were used, they would
ideally be
time-shifted by 0.25 milli-seconds. In general, where N enabling signals are
used and the
expected period of anyone signal is T, then each control signal should be time
shifted by
approximately T/N.
Protection signals 505 can be generated based on a variety of battery
conditions sensed
by protection means including, but not limited to, low voltage, high voltage,
temperature
or high current. The protection signal will allow a given battery to ignore
the enable
signal 506 that would normally cause power to be taken from the battery. The
energy
storage means 502 may optionally output a digital disable signal that would
cause the
charging cycle to terminate prematurely. This signal, which is not shown on
Figure 5,
could be generated in the case of excess heat, current, voltage or magnetic
flux. Such a
signal can also protect the system from a short-circuit on the output 204.
It can also be seen from Figure 5 and Figure 6 that any battery or any
combination of
batteries may be removed from the system at any time. Provided the energy
storage
means 502 is of sufficient size that the remaining energy storage means can
power the
system, then the system will continue to function normally.
11
CA 02643684 2008-10-28
Although the description above contains much specificity, these should not be
construed
as limiting the scope of the invention but as merely providing illustrations
of some of the
presently preferred embodiments of this invention. Any digital gates or
signals may be
easily redefined such that they perform similar functions as inverse logic or
using
alternate gates or logic topology. Logic, analog detection and control means
may be
implemented using integrated circuitry, microprocessor control, software and
wireless
control.
Thus the scope of the invention should be determined by the appended claims
and their
legal equivalents rather than by the examples given.
12
