Automatic selection of uninterruptible power supply model. How to calculate UPS battery life

Published by the author - , - January 29, 2014

For simplicity, we have made calculation calculators:

Now let’s present the calculation algorithm:

1) Determine the total load power and constant discharge current.

2) We calculate the required battery capacity for a given autonomy.

3) Determine the type of battery

Example

Given: two LED strips with a power of 10W each and operating on 12V. Required autonomy: 10h. Service life: one year with daily use. Operating conditions: constant room temperature 20 degrees.

Find: minimum acceptable and optimal batteries for solving the problem.

Solution

1) Total power W=10W*2=20W. Constant discharge current: I=20/12=1.67A. For accurate calculations, it is advisable to measure the current consumption using a multimeter.

2) To determine the required capacity, you should go through the following points:

A) In order to support the load at such a discharge current, it is necessary to determine the minimum calculated battery capacity: 1.67 * 10 = 16.7 Ah.

b) It must be borne in mind that the capacity of rechargeable batteries is indicated by manufacturers based on a certain discharge time. Usually it is 10 hours. But some manufacturers indicate 20 hours. Here we can get help with the battery, which you can get on our website. Let's look at the specification:

In our case, the operating time from the battery is 10 hours, which means we can consider the capacity equal to the nominal one. However, if the task requires 5 hours, then you need to make an allowance for the fact that with such a discharge time the battery capacity will be lower (we multiply the discharge current by the hours - 4.8A * 5h = 24Ah instead of 28).

In the problem we can see that the planned number of cycles is 365. The approximate maximum discharge depth in our case is about 57%. It is advisable to take it with a reserve; we will count on a 50% discharge (real operating conditions differ from ideal laboratory conditions).

Thus, we introduce a correction of 0.5: 16.7/0.8 = 33.4 Ah.

G) If we are dealing with an operating temperature other than the optimal one (25 degrees), it is necessary to enter a correction factor, which we can also take from the specification:

So at a temperature of 10 degrees you should enter a coefficient of 0.9, i.e. another +10% to the calculated capacity.

3) If we need long discharge modes, we should pay attention to the AGM battery series from popular manufacturers on the Russian market:

The battery has a Delta series
At CSB -

Determine the purpose of using the UPS and its primary purpose. An important indicator is the frequency of use and type of connected load. In the case when a UPS is being selected for home, household use with rare power outages, a line-interactive or backup type can be considered. In the case of commercial or industrial applications, an on-line type device will be required. Please note that these are general recommendations; to determine the correct choice of UPS class, call the sales department of guaranteed power supply systems (+380 44 383 3663).

3. Selection of UPS by characteristics

Using a calculator, select a UPS based on the specified parameters. The calculator database contains more than 16,000 models of uninterruptible power supplies and voltage inverters. The choice of UPS is based on an updated database, where most foreign brands are present: General Electric, INVT, Riello UPS, Socomec, Borri, Emerson, Eaton, APC, Legrand, Voltitronic, Ippon, as well as Ukrainian manufacturers: Reserve, Volter, SinPro, Integral, Phantom and others. The database also contains autonomous and hybrid voltage inverters of line-interactive and backup type Victron Energy, Power Star, Stark Country, MeanWell, TBS Electronics and others.

Attention! The database does not contain budget models of Ukrainian and Chinese production, which do not meet the stated technical characteristics, do not provide adequate battery charge and belong to an unreliable class of devices.

As of 2017, we can highlight some proven brands that have won a significant share of the Ukrainian market. Conventionally, UPSs and inverters can be divided into several segments based on cost, but you should not assume that the cheaper the product, the more often it will fail. As the experience of using our clients shows, even inexpensive UPSs can work no less efficiently, because the main thing is to make the right technical specifications and select a suitable model in full accordance with them.

A little theory

To calculate the operating time of an uninterruptible power supply (UPS) with any load, you need to know the battery capacity, which is expressed in ampere-hours (A*h). However, in the characteristics of a UPS they usually write not ampere-hours, but volt-amperes (V*A), that is, they write power. But this is not just power, but the ideal power output invented by marketers. The key word here is "ideal". That is, one that cannot exist in the real world. Let's denote it as Pideal.

More honest manufacturers indicate the effective power, which is traditionally expressed in watts. Let's denote it as Effective . Effective power is obtained from ideal power by multiplying by power factor:

Effective = k * Pideal

What is the power factor? k ? An inverter is installed at the output of the UPS, which converts the 12V supplied by the battery into 220V required to power the connected devices. Since the output current is alternating, the power loss is 1/sqrt(2)=0.70. Plus, we exclude from this power the power supply of the UPS circuit itself, and we get a coefficient approximately equal to 0.6.

For example, a regular office uninterruptible power supply APC Smart UPS 500 has a power of 500 VA. This one perfect power that the battery installed inside the UPS can provide. Efficient power, according to our formula and coefficient, will be only 0.6 of the ideal, i.e. 300 W.

Now the question. Why did we first write volt-amperes, and then began to write watts? Both are units of power. Traditionally, ideal power is written in volt-amperes, and effective power is written in watts. But these are quantities of the same dimension.

Calculating device operating time

Now let's understand how to calculate the operating time of a device powered by a UPS. For example, we have a managed Cisco router that consumes 50 watts. What does it mean to consume 50 W? This means that per hour he will spend 50 W of power for his work. That is, in fact, we should write 50 W/h. Let us denote this quantity as Dpower (power demand - power consumption).

Our UPS has an effective power reserve of only 300 W. This means that if the equipment consumes 50 Wh/h, then our UPS will be enough for:

300 W / 50 W/h = 6 h

That is, the formula for calculating time will be like this:

T = Effective / Dpower

That is, if Dpower will be in the dimension W/h, then the time will be in hours.

And finally, a little nonsense

Looking at the dimensions of power (volt*ampere), we recall the formula for electrical power from a school physics course:

P = U*I

Where:

P is the battery power, expressed in volt-amperes (V*A),
V is the battery voltage, expressed in volts (V),
I is the current generated by the battery, expressed in amperes (A).

Now, knowing that uninterruptible power supplies usually contain batteries with a voltage of 12V, we can find out the current strength that the battery can provide:

I = P/U = 500/12 = 41.6 A

Oh, wow, 41.6 A! What kind of current is this? This is normal current. It’s just a short circuit current when there is no resistance, and a current calculated based on ideal power. But you won’t short-circuit the battery, you will connect the load to the UPS.

March 29, 2016

Accurately calculating battery life using mathematical calculations is not a trivial task. In this regard, we simplified the task by implementing the calculation algorithm in calculators:

However, let's look at approaches to determining battery life.

1) Simple formula

T = E U / P

E - battery capacity in Ah
U - voltage
P - load power in W.

This is a highly simplified formula that gives a very approximate result for discharges in the range of 5-15 hours. Suitable for quickly estimating the autonomy time in your head. The algorithm does not take into account the decrease in battery energy output during short discharges and the increase during long discharges, as well as various coefficients.

There is an improved formula with coefficients:

T = Uab * Sak * K * h * Kr * Kg / Pnagr

T – battery life of the uninterruptible power supply, h;
Uab – battery voltage, V;
Sak battery capacity, Ah;
K – number of batteries in the circuit;
h – converter efficiency (h=0.75-0.9), often varies depending on the load;
Kr – discharge depth coefficient 0.8–0.9 (80%-90%), should be considered 80%;
Kg – coefficient of available capacity (depends on the discharge mode and temperature, see battery characteristics)
Rload – load power.

This algorithm gives relatively accurate results, but for long-term discharges of 1 hour or more. At short discharges, the results can be highly distorted due to the nonlinear discharge function of lead-acid batteries. We used a similar method in the article.

2) Peukert formula

T=Cp/I^n

T – time in hours
Cp – Peckert capacity (battery capacity when discharged with a current of 1A)
I – discharge current
n – Peukert exponent

The Peukert exponent is sometimes indicated in the battery characteristics, and it is calculated based on the battery’s C-rating data (capacity at different discharge times). Peukert capacity is calculated by the formula – Ср=R(C/R)^n (R is the rating in hours corresponding to this capacity, for example, 10).

Our calculators are based on this formula, taking into account the efficiency of inverters and the depth of discharge. They calculate the autonomy time with high accuracy for both short and long discharges.

3) Calculation using tables from battery specifications

Step 1. Calculation of total power into battery load power

Rakb = (Pload*cos(φ)*Knagr)/efficiencyinv

Pload – power in kVA
cos(φ) – power factor characteristic (load characteristic)
Knagr – UPS load level
Efficiency inv – inverter efficiency

For example, let's take a 120 kVA UPS operating at 70% load with a power factor of 0.8:

Rakb= (120000*0.8*0.7)/0.94=71,489W - It is this load that will fall on the entire battery bank when the UPS is powered from the battery.

Step 2. Calculate the load on one battery

Let's recalculate the load on one battery. As a rule, in large UPS batteries are connected in series in a quantity of 32-40 pcs. To calculate the load on one battery with 40 batteries:

71,489W/40=1,788W.

The battery data sheet usually indicates the power per cell (Pel), of which there are 6 pieces. in a 12V battery. Hence:

Rel = 1788/6 = 298W.

Step 3. Study of battery discharge tables and selection.

In the article we looked at the subtypes of batteries in the context of different intended uses. One of the basic characteristics is energy output, i.e. how much power the battery is capable of delivering in a certain time.

Let's look at the discharge tables of 100Ah Delta batteries of two different series.

Delta DTM 12100 l:

Delta HRL 12100:

Let us remember that our load on the element is 298W. Discharge depth – 10.8V or 1.80V per element. Thus, from these tables, we can conclude that DTM 12100 l will support the load for about 13.8 minutes (can be calculated proportionally, distortion is minimal), Delta HRL 12100 - 16.3 minutes. order difference 15% . By the way, the price difference is approximately the same.

4) Carrying out real discharges

Of course, the ideal is to conduct actual bit tests. It must be taken into account that batteries reach their maximum capacity by the 10th charge-discharge cycle.

Let's remember some physics

When estimating the power consumed by a load, the total power should be taken into account. Apparent power (VA unit of measurement is volt-amperes) is the entire power consumed by an electrical appliance. It consists of active (unit of measurement "W" - Watt) and reactive (unit of measurement VAR - volt-ampere reactive) power components. Electricity consumers often have both active and reactive components.

. With this type of load, all consumed energy is converted into heat. For a number of devices this component is the main one. These include, for example, electric stoves, lighting lamps, electric heaters, irons, heating elements, etc.

Reactive Loads . Almost everything else. They can be inductive or capacitive in nature. A typical representative of an electrical device that has an inductive load component is an electric motor. Apparent power (P) and active power (Pa) are related by the coefficient cosФ.

Ra = cosФ x P

What is the methodology for calculating the power of electrical consumers?

In order to make the optimal choice of UPS model based on the required power criterion, you need to calculate the total power consumed by your load. Load, in this case, means all electrical appliances located in your home (office, apartment, industrial premises) that are subject to protection.

The power consumed by a specific device can best be determined from the data sheet or operating instructions for this product. Sometimes the power consumption and cosF coefficient are indicated on the back wall of the device or device. It should be noted that the power value in documents for different devices can be indicated either in watts or in volt-amperes. In order to avoid errors when calculating the power of devices, we summarize separately for each unit of measurement in two columns.

We will list all electrical consumers subject to protection;
Let's sum up their powers as indicated above;
Let's bring the results obtained to one unit of power measurement (preferably in volt-amperes). For this:
If the active power and cosF coefficient are indicated in the passport, then it is easy to recalculate it into full power. To do this, the active power in "W" must be divided by cosФ. For example, if the product says that the active power is 700 W and cosФ = 0.7, then this means that the total power consumed will be equal to 700/0.7 = 1000 VA. If cosФ is not specified, then for an approximate calculation we will take it equal to 0.7.

The power calculated in this way should be added to the sum of the powers in the other column (summed up in VA).

Note: for electrical appliances that have only an active load, the cosФ coefficient is taken equal to 1.

Another extremely important point should be taken into account - inrush currents. Any electric motor (compressor) at the moment of switching on consumes several times more energy than in nominal mode. In the case when the load includes an electric motor (for example: a submersible pump, refrigerator, drill), its rated power consumption must be multiplied by at least 3 (preferably 5) in order to avoid overloading the stabilizer or UPS when the device is turned on. Make these adjustments to your calculations.

So, the power has been calculated.

However, let's take two more points into account.

There are practically no cases in life when absolutely the entire load is working at the same time. In fact, if you are welcoming guests, then it is unlikely that the laundry is washed at this time, the lighting is not turned on during the day, etc. In practice, there is such a thing as the “simultaneous switching coefficient”. Thus, the calculated value can be reduced (i.e., multiplied by approximately a factor of 0.3-0.5).
On the other hand, it is unacceptable for it to work in full load mode. To create a “gentle” operating mode, it is advisable to increase the power obtained as a result of previous calculations by approximately 10-15%. By doing this, you increase the service life of the equipment, increase reliability and create a power reserve for connecting new equipment.

The required number has been found. Now, based on specific examples, let's choose a UPS.

To facilitate the task of determining power, you can provide a table with approximate data on the electricity consumption of household appliances.