This case is an extension of the previous case of online UPS (click here) where a UPS was used to supply the load when the main source failed. In this case, a UPS is considered to be feeding a load. Any UPS will have a load limit beyond which it will enter a current saturation mode and the output voltage will fall. Therefore if the load connected to the UPS is excessive or additional load switched on to the UPS output results in the UPS entering the current saturation mode, the output of the UPS will fall and this will affect all the loads connected at the output of the UPS. This case will examine a possible solution using a storage system connected at the output of the UPS through a dc-ac converter. The storage system is considered a simple high voltage battery with a charge and discharge cycle that depends on the current drawn or supplied by the battery. The objective of controlling the battery is to monitor the output voltage of the UPS and if this falls, consider it to be an indication of overloading. In this case the battery discharges through its interfacing dc-ac converter operating in current control mode. The additional current supplied by the battery will relieve the UPS from supplying all the current demanded by the load and the UPS output voltage is restored. However, the battery discharges and when the voltage drops to a specified level, the battery stops supplying current to prevent a discharge that can lessen its life. At this point, if the overloading persists, the UPS output voltage drops. When the overloading is withdrawn by load shedding to prevent malfunctioning of critical load due to low voltage, the UPS output voltage is restored. At this point, a healthy system state is attained. The battery which is in a discharged state will now charge by drawing a nominal current from the UPS. This operation will be shown by simulation results.
Let us examine the start-up of the system. The figure below shows the output voltages of the UPS as it slowly regulates the output voltage to 208 V line-to-line. Until 0.1s, the UPS is the only converter and the storage system is in a cut-off mode until it synchronizes with the UPS output voltage through its PLL. At 0.1s, the the control of the storage system starts as it has a dedicated control system which is now locked on and ready.
The voltage of the battery starts at 450V. The figure below shows the currents exchanged by the battery. The battery exchanges these currents with the UPS until its voltage exceeds 450V. When the battery voltage exceeds 450V, it goes into a cut-off mode to avoid excessive charge to the battery. The currents become zero. When examining the UPS voltages above, it can be seen that the disturbance in the UPS voltages dies down a few cycles after 0.1s.
The figure below shows the voltage of the battery. It starts at 450V. When the battery converter is in cut-off mode and receives no switching pulses, the drop in the battery voltage is negligible. But this negligible drop below 450V is sufficient to trigger the charge control at 0.1s. The battery voltage rises above 450V and now the converter returns to cut-off mode. Typically this nuisance operation can be avoided, but even if it is not, there will be a short pulse of current after several seconds to replenish the charge.
Now to the load. At the beginning and until 0.3s, only a single three-phase R-L load is connected and this does not overload the UPS. The UPS voltage remains well regulated. The figure below shows the total current supplied by the UPS.
Now to examine the transient that occurs at 0.3s. At 0.3s, another large load is connected in parallel to the existing three-phase load at the output of the UPS. The figure below shows the load current and how it spikes at 0.3s. The load current falls soon a cycle after the spike and the reason is apparent when comparing with the UPS output voltage.
Below is the plot of the UPS output voltages at the transient. As can be seen, when the load increases, the UPS goes into a current saturation mode, and the output voltage falls.
This is when the storage system kicks in. The battery system monitors the UPS output voltage. If assuming that the source of the UPS is a constant dc voltage and that any drop in the output of the UPS is due to excessive load and not failure of the source, a drop in UPS voltage becomes the sign of overloading. The battery control enables the interfacing dc-ac inverter and supplies a current to the system. This can be shown in the plot below.
The plot below shows the output voltages of the UPS due to the action of the battery system. As can be seen, the UPS voltages rise back to the their usual voltages. This is due to the current supplied by the battery system which relieves the UPS from supplying the entire load current.
The currents supplied by the UPS are shown in the plot below. As can be seen, the current spikes at 0.3s due to the overload but drops once the battery begins to supply a part of the load current.
The battery will discharge in the above manner until two possible conditions occur. First, the battery voltage falls below 390V at which time the battery control cuts-off the pulses to the interfacing inverter. Second, the overload is removed and the total current demanded by the load drops below a threshold. In this simulation case, the first case is considered where the overload lasts for a period long enough for the battery to discharge and go into cut-off mode. The plot below shows the discharge of the battery along with the battery current until cut-off occurs.
The plot below shows how the battery voltage drops when it discharges.
When the battery reaches cut-off mode, the UPS voltage again collapses due to overload. This can be seen in the next two plots - the first being the UPS voltages and the second being the load current.
At this point, either the system could continue until the overload is removed. Or a load shedding control could switch off excessive load to allow the UPS to recover. Let us assume that such a load shedding scheme disconnects the excessive load at 0.8s. Since, the load current drops, the UPS voltage recovers as shown below.
The load currents are also restored to their values before the overload.
Once the UPS voltages are restored and the total load current is back to a normal value, it is now time for the battery to charge and restore its voltage. The battery control restores pulses to the interfacing inverter and now draws a nominal current from the UPS.
The battery voltage gradually rises as it draws only a nominal power from the UPS. To give an indication of the lower charge rate with respect to the discharge rate, the battery voltage is plotted for the entire duration of the simulation which is 2s.
This simulation has described the basic concept of how a battery can be interfaced to a system and controlled to provide relief from overloading. Alternatively, it can also be used to size the battery as this simulation shows how the battery discharges to a level where it has to cut-off despite the overload still persisting. In such a case, a larger battery would be needed to ride through the overload without performing load-shedding. Another scenario would be to use this simulation to design a load shedding algorithm to ensure that the voltage supplied to the load will never go out of bounds.