BATTERY DISCONNECT METHOD AND SYSTEM
TECHNICAL FIELD
The present invention relates to a battery disconnect method and system and also to a method and apparatus for determining a disconnect reference voltage.
BACKGROUND
Batteries are often used to provide back-up of mains supplies. Important equipment, such as telecommunication equipment, computers, hospital equipment etc. are often powered by special power systems that can provide electrical power even if the mains supply fails. Typically such power systems are provided with (rechargeable) batteries that supply the necessary electrical power in case the mains supply fails. In general such batteries should not be overdischarged. In particular the most common battery, the lead-acid battery, should not be discharged beyond a certain safe limit. Overdischarge of such batteries may cause the battery to become non-functional and may even cause external danger. In order to avoid this situation, the loads must be disconnected from the battery when the limit has been reached.
A complication is that the safe limit or disconnect voltage depends on the rate at which the battery is discharged. For example, if a battery is di en- sioned for a backup time of 10 minutes, it is safe to disconnect the loads at a battery output voltage as low as 1.6 V per battery cell. On the other and, if the battery is intended for a backup time of 200 hours or more, the disconnect voltage has to be set at 1.9 V per cell or more. Thus, a low discharge rate requires a higher disconnect voltage than a high discharge rate. This means that if a single safe limit acceptable for all situations is to be used, it must be set high enough to cope with low discharge rates. However, this would be a non-optimal utilization of the battery in cases where it is known
that the discharge rate will be high. A solution to this dilemma described in [1] sets the safe disconnect voltage to a value that is inversely proportional to the measured discharge current. However, a major drawback of this method is that it requires measurement of the discharge current, which is a complicated and undesirable feature. Another drawback of the known method is that the disconnect control system requires configuration for each battery capacity (Ah) due to the discharge current based disconnect voltage calculation method. These drawbacks have meant that the method has not been widely used.
Reference [2] describes a fast charge termination method, in which the first time derivative of the battery output voltage is used to determine a charge teraiination point. F ιrtherrnore, the second time derivative of the battery output voltage is used to deterπiine the concavity of the battery output voltage in order to eliminate erroneous voltage peaks near the start of a charge cycle.
SUMMARY
An object of the present invention is to provide a configuration- free or essentially configuration-free battery disconnect method and system that are not based on current measurements.
Another object is a method and apparatus for determining a battery discon- nect reference voltage that are not based on current measurements.
These objects are achieved in accordance with the attached claims.
Briefly, the present invention is based on the discovery that the output voltage of a battery has a parabolic time variation for each specific discharge rate. This fact may be used to calculate a measure that is independent of discharge rate from the first and second time derivatives of the battery output voltage. This measure is used to determine a dynamic disconnect
reference voltage. Since the disconnect reference voltage is based on a measure that is independent of the battery capacity, this disconnect system has the advantage of being configuration-free.
BRIEF DESCRIPTION OF THE DRAWINGS
The invention, together with further objects and advantages thereof, may best be understood by making reference to the following description taken together with the accompanying drawings, in which: Fig. 1 is a block diagram of an exemplary embodiment of a battery disconnect system in accordance with the present invention;
Fig. 2 is a time diagram illustrating different discharge curves for a battery;
Fig. 3 is a block diagram of an exemplary embodiment of a battery disconnect unit suitable for the battery disconnect system in fig. 1; and
Fig. 4 is a flow chart of an exemplary embodiment of the battery disconnect method in accordance with the present invention.
DETAILED DESCRIPTION
Fig. 1 is a block diagram of an exemplary embodiment of a battery disconnect system in accordance with the present invention. A mains supply 10 with a rectifier is connected to one or several loads 12. If the mains supply fails, a backup battery 14 will temporarily replace mains supply 10 over a normally closed switch 16. A disconnect control unit 18, which will be further described with reference to fig. 3, measures the battery output voltage U over an input line 20 and determines a disconnect reference voltage that is compared to this voltage U. If battery output voltage U falls below this disconnect reference voltage, unit 18 will disconnect loads 12 from battery 14 by opening switch 16 over a control line 22 (In the following description switches 16A-C and control lines 22A-C are temporarily ignored; switches 16A-C may simply be consid-
ered as permanently closed). Suitable switches are contactors, circuit breakers or semi-conductor switches.
Fig. 2 is a time diagram illustrating different discharge curves for a battery. It is helpful to study these discharge curves before disconnect control unit 18 is described in detail. Fig. 2 illustrates 3 different discharge curves, namely a fast discharge curve Cl, a medium fast discharge curve C2 and a slow discharge curve C3. The discharge rate is assumed to be constant in each curve. As can be seen from this family of discharge curves, the output voltage U drops more rapidly for higher discharge rates. However, surprisingly it has been found that the entire family of curves may accurately be represented by a set of parabolas. Thus, a general discharge curve may be represented by the parabola:
U(t) = At2 + UQ
where A, Uo are constants. Here Uo represents the common starting point for all curves (the output voltage of a fully charged battery), while A represents the "width" of each parabola. Differentiating this equation twice gives:
U t) = 2At
U"(t) = 2A
Eliminating A and solving for t gives:
,=EX1_
U"(t)
Thus, from the parabolic model described above it is possible to obtain a measure t, called the effective discharge rate tEFF below, that is independent of the actual discharge curve ("width" coefficient A). The parameter tEFF relates to the extent a battery has been discharged at the present discharge rate. If there
are changes in the load during battery operation, the discharge curve will also change. For instance, if the load increases from a low value to a much higher value, the discharge curve will be "depressed" and have a higher curvature. In this case tEF will in fact decrease, although the total elapsed time has increased. Thus, tEFF may be considered as a measure of discharge rate.
The recommendations for safe disconnect voltages vary between different lead- acid battery manufacturers, but the table below; gives an overview of the recommended values.
Standard curve fitting techniques may be used to have these recommendations in a more convenient form. A logarithmic relation has been found to be suitable (However, the relation between discharge rate and disconnect voltage is not particularly critical, many other models are also possible.). Thus, the disconnect reference voltage UREF may be expressed in accordance with the model:
UREF = B - log( EFF) + C XT tEFF - U"
where B and C are constants. This disconnect reference voltage curve is illustrated by the dashed line in fig. 2. If "log" represents the natural logarithm, suitable values for B and C are, for example, 0.038 and 1.70, respectively. These values are for a single lead-acid battery cell. For several cells UREF is multiplied by the number of cells.
Fig. 3 is a block diagram of an exemplary embodiment of a battery disconnect unit suitable for the battery disconnect system in fig. 1. The battery output voltage U is forwarded to a sampler and A/D converter 30 on input line 20. The reason for the sampling and A/D conversion is that today most batteries for backup are already provided with computational power in the form of a micro processor. In the illustrated embodiment this micro processor may be used to perform the essential functions of disconnect control unit 18. Typically the battery voltage U is sampled at a sampling frequency of 0.1 - 100 Hz. The digitized voltage signal is forwarded to a differentiator 32 that forms a measure of the time derivative U' of output voltage U, for example the difference between two neighboring samples. The time derivative U' is forwarded to another differentiator 34 that forms a measure of the second time derivative U" of output voltage U, for example the difference between two neighboring samples of the first derivative U\ An inverter 36 inverts U", and the result is multiplied by U' in a multiplier 38. The product from multiplier 38 is forwarded to a block 40 forming the logarithm of its input signal. Another multiplier 42 multiplies this logarithm by a constant B from a memory cell 44. An adder 46 adds the product to a further constant C stored in a memory cell 48. The resulting disconnect reference signal UREF is compared to the current battery output signal U in a comparator 50. "When the battery output voltage
U falls below the disconnect reference voltage UREF, control signal line 22 will trigger switch 16 in fig. 1 to disconnect battery 14 from load 12.
Fig. 4 is a flow chart of an exemplary embodiment of the battery disconnect method in accordance with the present invention. Step SI performs an A/D conversion of battery output voltage U. Step S2 differentiates the voltage with respect to time. Step S3 performs a second time differentiation. Step S4 determines tEFF. Step S5 forms UREF from tEFF. Step S6 determines whether U>UREF. If not, step S7 disconnects the loads from the battery. Otherwise the procedure returns to step SI to digitize the next sample of battery output voltage U.
In the exemplary embodiments of the invention described above, all loads are simultaneously disconnected from the battery. However, sometimes it would be preferable to disconnect the loads one by one or to disconnect different groups of loads separately. One situation in which this is appropriate is when the loads are of different importance ("priority loads"). In this case it would be desirable to have the high priority loads connected as long as possible. Quite often such high priority loads are low power loads, which has been a problem with previously known battery backup systems, since low power loads would require a high disconnect voltage, which would make poor use of the battery. With the present invention, however, it is possible to meet both objectives. The disconnect reference voltage UREF will still be the proper disconnect voltage immediately after the change^ of discharge curve. Thus, by providing different loads with different switches and by associating each switch with a different constant C, it is possible to individually disconnect loads to guarantee that high priority loads will be the last to be switched off. This embodiment is illustrated by switches 16A-C and control lines 22A-C in fig. 1.
Another possible modification is the sampling procedure. The described embodiment has a constant sampling rate. An alternative is to let the battery output voltage level trigger when samples of U' and U" should be determined.
Furthermore, the present invention has been described with reference to lead-acid batteries. However, the same principles may also be applied to other batteries with essentially parabolic discharge curves. Examples are Lithium ion batteries and Manganese dioxide - Zinc batteries.
It will be understood by those skilled in the art that various modifications and changes may be made to the present invention without departure from the scope thereof, which is defined by the appended claims.
REFERENCES
[1] U.S. Patent No. 4 086 525 (Ibsen et al.)
[2] U.S. Patent No. 5 563 496 (McClure)