[01] The energy stored in a capacitor is nothing but the electric potential energy and is related to the voltage and charge on the capacitor. If the capacitance of a conductor is C, then it is initially uncharged and it acquires a potential difference V when connected to a battery. If q is the charge on the plate at that time, then q = CV. The concept here is that instead of increasing C to increase energy density, it should be more cost effective, although potentially less energy dense to increase V. But, for stationary and secondary sources as uses, cost effectiveness should be more relevant, therefore increasing the voltage is the part of the solution proposed by the present invention.

[02] The challenges with the use of high voltage capacitor or capacitor bank are related with greatly varying charges when charging and, therefore, throughout greatly varying impedances, and to discharge at a usable lower voltage. Therefore, the invention, as the solution, must comprise a charging device, a high voltage capacitor or capacitor bank and a discharging device. Therefore, the invention relates to a stationary energy storage system device, comprising: a. a semi-constant power supply that supplies semi-constant power throughout the impedance range of the charging capacitors. b. High voltage capacitors. c. A buck (or buck-boost) converter comprised by at least one switching component, at least one capacitor and at least one inductor.

[03] Complimentary components being: a. A pre-charge system. b. A mono or triphasic 50-70Hz 110-440V inverter with a sinusoidal output. [04] Object of the current invention is to provide for a use of cheap high voltage capacitors as a stationary storage system. The advantage is the cost effectiveness of cheap such solutions that can be engineered from simple components such as aluminum, activated carbon and high voltage dielectrics. This type of capacitor can have a relatively large energy density and a huge power density, but with the potential disadvantage of being extremely hard to charge and discharge.

[05] Considering the main use of this type of energy storage can be stationary or secondary power supplies, they can be insulated or buried, so safety should not be a risk.

[06] Charging is challenging, because a voltage source would be overly demanded when the capacitor charge is low and would have to be prepared to deliver hugely different voltages throughout changing impedances to keep power consumption at the charging side minimally constant.

[07] Discharging is challenging because the voltage of the capacitors is much higher than the regular consumption voltage levels, so a voltage step down is required before the final inverter.

[08] Traditional systems that have fixed voltage DC, AC or greatly varying voltages with fixed waveforms with high voltage peaks, but when the capacitor bank to be charged can withstand high voltages, it makes sense to charge them completely, which means these traditional systems will deliver a lot of power when the charge is low and just a little when the charge is high, demanding a lot from the components and from the power source.

[09] The solution for the charging is a semi-constant power supply that supplies semi-constant power throughout the impedance range of the charging capacitors comprised of at least one electronic converter composed of, at least, the following components: a. A power supply. b. A frequency converter such as an inverter with a controlled PWM signal (such as the one in Fig. 9.) that ensures that the voltage output does not have peak amplitude voltages of more than 5kV of parasitic waveforms, and that the voltage output is controlled to ensure that the power output is constant throughout the usual operational load range. c. a transformer. d. a voltage multiplier or rectifier composed of diodes and capacitors.

[10] The impedance faced by the charging is basically composed of the capacitance of the capacitor bank and any resistances or inductors that may compose a pre-charge circuit. This may vary greatly because of the varying charge on the capacitors, therefore a semi-constant power supply that supplies semi-constant power throughout the impedance range of the charging capacitors is necessary not to overstress the power source.

[11] The DC voltage usual values should be in the range of 1 - 1 OOkV, which is usually necessary to ensure constant power delivery in the commonly found load range of a regular operation. The peak DC voltage is defined as the maximum DC voltage.

[12] To avoid damaging the dielectric components of the capacitors, it is relevant to control any voltage peaks. This means that the ideally, the processed electrical energy output should not comprise a waveform with repeating shapes, but a DC that varies its voltage exclusively according to the load it faces to ensure constant power delivery, up to a set maximum that does not exceed the maximum the capacitors can withstand. Given the cycles of charge and discharge of the capacitors of the voltage multiplier, when the load is low, parasite voltage peaks may be seen, but should be kept as low as possible and preferably below IkV. [13] Parasitic peak voltages may arise as the charge of the capacitors decrease faster than the input from the secondary of the transformer can charge. This effect is especially high when either loads are too low or switching frequencies are too low. When loads are too low, the capacitors discharge too fast and when switching frequencies are too low, the PWM or similar controls cannot act fast enough to counter act the discharge. Typical switching frequencies are above 1kHz and below 10MHz.

[14] The parasitic peak voltage is defined as the peak amplitude of the repeating units of the parasitic waveform (Fig. 1) which superimposes the variable DC voltage. Peak-to-peak voltage (Fig. 2) is greater than the peak voltage and typically twice that of a sinusoidal waveform. The parasitic waveform is the repeating unit defined as a unit that repeats with substantially the same form, e.g. it may comprise waveforms of substantial the same shape including when the amplitude and/or duty cycle or period is adjusted for control of the processed electrical energy.

[15] The addition of an inductor and/or inductive/capacitive filter may allow for the inverter’s IGBT to work with resonant switching, reducing its losses and increases the converter’s efficiency. The IGBT’s in this case is tuned at the resonance between the external inductor and the total capacitance reflected to the transformer’s primary or directly to the voltage multiplier, considering the effects of the variable load and the voltage multiplier.

[16] The patent application EP3557750 Al object of invention contains a frequency converter, a transformer and a capacitive voltage multiplier composed of diodes and capacitors. To ensure semi-constant power, it uses the impedance matching implicit of the voltage multiplier, which happens in a self-adjustable way, without the necessity of a control strategy implementation. This happens to a certain degree because when the resistive load tends to a low value (short-circuit situation), the voltage multiplier presents a series impedance reflected to the primary that, associated with the external inductor of the filter, protects the transformer against high short-circuit currents. When the load tends to a high value (open circuit situation), all the capacitors of the voltage multiplier are charged, increasing the secondary voltage peak, but still limiting it to a maximum value equals the multiplier stage. In other words, when the load is higher than the load that delivers maximum power, the power diminishes because the converter cannot increase the voltage enough to ensure constant power, and when the load is lower than the load that deliver maximum power, the internal impedance reflected increases, lowering the power delivered. A typical power curve (y axis in watts - w) against the load (x axis in kilo ohms - kQ) can be found in Fig. 3. The actual graph with the power curve of the invention shown in that patent can be found in Fig. 4.

[17] Although it has some degree of power control, it is semiconstant, given the nature of the power output curve of a regular voltage multiplier. To improve upon this invention, a multiple setting of different values for the voltage multiplier can be constructed. For instance, paralleling multiple voltage multipliers with different peaks may provide a power curve with multiple peaks, where the power output is stable at a much wider load range. A combination of two and three of those can be seen in Fig. 5 and Fig. 6, where two and three peaks can be seen, respectively, and where a growing stable area of semi-constant power can be found.

[18] Now, if a pulse width modulation (PWM), pulse density modulation (PDM) or any other similar control is used as a feedback loop is incorporated to the improved design, ensuring that the capacitor charge will limit the output voltage to the exact degree that outputs the desired voltage (i.e. 250w in this example), the converter may find a large load range where it can function with stable constant power, as shown in Fig. 7. [19] For that, the needed information for the feedback loop is the output power. If the output power deviates upwards, the PWM (or similar) decreases the duty cycle, if it deviates downwards, it increases the duty cycle.

[20] Duty cycle is the amount of time a digital signal is in the “active” state relative to the period of the signal. Duty cycle is usually given as a percentage. For example, a perfect square wave with equal high time and low time has a duty cycle of 50%. Fig 8. The duty cycle controls the gates of the inverter (usually IGBTs are used).

[21] The duty cycle adjustments will provide more power keeping the charge of the capacitors of the voltage multiplier at a lower voltage state when the load is lower and keeping the charge of the capacitors of the voltage multiplier at a higher voltage state when the load is higher. This means that with the combination of an architecture of voltage multipliers with multiple voltage peaks throughout a continuum of loads with a PWM control that ensure power delivery never goes above a certain value, it is possible to provide constant power throughout a large load band, and through a dynamically varying load. According to Ohm’s law, the relationship between power, voltage and load should follow Fig. 10 for a constant power source.

[22] A very simple embodiment of such electronic architecture with only two multiplier arrays connected using a mid-tap reference of the transformer can be found in Fig. 11. For optimal usage the person skilled may want more complex arrays and combinations of multipliers with multiple peaks to ensure a semi-constant power delivery throughout a larger band of resistances, reducing the usage of the PWM, specially at lower duty cycles, which are potentially more lossy and less energy efficient. It considers the output of a DC input power source as VDC1 and VDC 2, switching inputs to control PWM or similar at SI and S2, higher frequency inputs at the transformer primary as HFA1 and HFB1, higher frequency outputs of the transformer secondary as HFA3 and HFB3, two parallel arrays of multipliers using a common middle capacitor MPY C2, the final outputs of the charging is a semi-constant power supply that supplies semi-constant power throughout the impedance range of the charging capacitors are set in the design as Ground and X.

[23] Although the charging system should deliver constant or at least semi-constant power, at very low impedances the current requirements may be to high, therefore a pre-charge circuit may be needed. A traditional embodiment can be found in Fig. 12.

[24] The high voltage capacitors can be any combination of existing capacitors in series or, more suited to this kind of use, capacitors made from cheaply available materials which may not offer very high energy density but can be very cost effective. An example would be capacitors made of aluminum foil as the conductive parallel plate electrodes, activated carbon with a conductive substance to increase area, and ceramic or plastic dielectrics.

[25] Due to have varying voltages with potentially high maximum, this system may not be suited to feed directly any loads of interest, such as regular 60Hz sinusoidal inverters that feed power outlets. For this reason, a voltage step-down converter such as a Buck is required. This voltage step-down converter must deliver constant or semi-constant voltage to the load of interest, in usable voltages such as 10-500V. This Buck should be composed of, at least, an inductor, a capacitor and a switching element such as an IGBT. A traditional embodiment can be found in Fig. 13.

[26] The main challenge is the voltage through which those components may be exposed to. The inductor is of an easier construction, the capacitors can be solved in the same manner as the capacitor bank, and the switching element can be a high voltage switching element or a combination of multiple switching elements in series with or without additional components to ensure no over-voltage or current applies to any single switching element. [27] The output of the Buck system can be used directly or fed to an inverter to produce usable electrical energy. Such traditional inverters can be widely found in the solar panel industry e.g.