How to Calculate Generator Size Using Starting Demand and Duty Cycles

Why Standard Nameplate Ratings Fail

A diesel generator nameplate shows rated kVA, kW, and voltage. It does not tell you what your specific facility will demand. Buying based on total connected load often results in a machine that is too large, wasting capital. Buying too small causes tripping, voltage sag, or fuel starvation during peak starts.

Generator load calculation is a mathematical process. It requires separating loads into categories, applying correction factors, and defining operating time. This guide walks through the actual arithmetic used by engineers and facility managers. The nameplate rating is a manufacturer’s guarantee under controlled conditions. Your site is never controlled. You have a specific mix of motors, heaters, and pumps. You have a specific ambient temperature. You have a specific startup sequence. The calculation must bridge the gap between the factory test bench and your plant floor.

Prerequisites: What Data You Need

Before touching a calculator, gather these documents:

• Single-line diagram showing all equipment on the generator circuit
• Nameplate data for every motor, heater, and compressor
• Load schedule listing kW or kVA per circuit
• Operating instructions for each machine (start type, run time, idle frequency)
• Site power quality report, if available

Without the single-line diagram, you cannot determine which loads run simultaneously. Without nameplate data, starting demand calculations are guesses. You need the exact voltage and current ratings. You need to know if the motor is a direct-on-line start or a variable frequency drive. If the single-line diagram is outdated, walk the plant. Tag every breaker. Record the circuit ID and the connected equipment. A missing diagram is the most common cause of a failed generator start.

Step 1: Identify All Connected Loads

List every piece of equipment that will connect to the generator. Group them by type: induction motors, permanent magnet motors, heaters, compressors, pumps, and lighting. For each item, record rated kW, voltage, power factor, and start method.

Reason: Different equipment classes draw different currents. A 50 kW induction motor starts at roughly 4 to 6 times full load current. A 50 kW heater starts at 1.0 times. Mixing these without separation distorts the calculation.

Do not group equipment by floor or room. Group it by electrical behavior. A 50 kW compressor motor and a 50 kW resistance heater both draw 50 kW at steady state, but they behave differently during the first few seconds of operation. The heater draws a high inrush current briefly, but it has no mechanical inertia. The compressor motor has heavy flywheels and stators that require a massive torque to overcome. You must know the power factor at full load. If the nameplate is missing, assume 0.85 for induction motors and 1.0 for resistive loads. If you are unsure, measure it with a power analyzer. Guessing power factor leads to kVA errors.

Step 2: Calculate Continuous Load

Sum the kW of all equipment that runs continuously during normal operation. This includes lighting, HVAC, process pumps, and steady-state production lines. Do not include motors that start and stop frequently, unless they are always running.

Reason: Continuous load sets the baseline. The generator must handle this indefinitely without overheating. Most manufacturers rate generators for 100% continuous power at 100% ambient temperature, but derating applies above 35°C.

Identify the base load. This is what runs 24/7. If you have a cooling tower fan that runs continuously, it is in this bucket. If you have a 100 kW conveyor belt that runs 12 hours a day, it is not a continuous load for the generator rating unless the generator is only used during those 12 hours. You must define the operating scenario. Is the generator for emergency power? If so, the continuous load is just the life safety systems. Is it for primary power? Then you include the production line.

Convert kW to kVA. If the power factor is 0.8, a 100 kW load is 125 kVA. The generator is rated in kVA. The engine and alternator are limited by kVA, not just kW. If you ignore power factor, you will undersize the alternator.

Step 3: Calculate Starting Demand

For each motor that will start while the generator is already carrying continuous load, estimate the starting current. Use the following general multipliers:

Motor Type	Starting Current Multiplier
DOL induction motor	4 to 6 times FLA
Soft starter induction motor	2 to 3 times FLA
VFD induction motor	1.5 to 2 times FLA
Permanent magnet motor	1.0 to 1.5 times FLA

Multiply each motor kW by its multiplier. Add this to the continuous load. Do not sum all starting demands. Only include motors that start simultaneously or within 2 seconds of each other.

Reason: Starting demand is transient. A single large motor start may spike the load, but the generator must handle it without voltage collapse. Simultaneous starts are the worst case.

Look at the control logic. If the plant startup sequence starts the main compressor, then 5 seconds later starts the auxiliary pump, the generator sees two distinct spikes. If they are far enough apart, the voltage recovers in between. You only calculate the worst-case simultaneous start. A direct-on-line start of a 50 kW motor at 5 times full load is a 250 kW spike. If you have three such motors starting at the same instant, that is 750 kW of starting demand. Add this to your continuous load. If the result exceeds the generator rating, you have a problem. You need a soft starter or a larger generator.

Step 4: Apply Diversity and Simultaneity Factors

Not every load operates at 100% capacity at the same instant. Apply a diversity factor to continuous loads. A typical range for mixed industrial facilities is 0.8 to 0.95. For commercial buildings, 0.7 to 0.9 is common.

Multiply the continuous load sum by your chosen diversity factor. Then add the simultaneous starting demand.

Reason: Diversity reduces over-sizing. If 20 HVAC units are installed but only 12 run at full capacity, calculating for all 20 wastes money.

Diversity is a judgment call. It is not a fixed number. Look at your load history. If you have a year of sub-meter data, use it. Find the 95th percentile load. That is your realistic continuous load. Without data, use the range above. A commercial office building with 100 HVAC units will never run all 100 units at full capacity at 3 PM on a Tuesday. Only 70 to 90 will be active. For a single production plant, the diversity might be higher because the process requires specific sequencing. If the process demands all units running, the diversity factor is 1.0.

Step 5: Analyze Duty Cycles

Duty cycles define how long the generator runs at each load level. Record:

• Load level (kW or kVA)
• Duration per cycle (minutes)
• Cycle period (total time from start to next start)
• Frequency of occurrence

For example, a 100 kW compressor runs for 15 minutes, stops for 15 minutes, repeating every 30 minutes. This is a 50% duty cycle.

Reason: Duty cycles affect cooling and fuel consumption. A generator running at 80% load for 24 hours needs more cooling capacity than one running at 80% for 2 hours daily.

This step determines thermal stress. The alternator and engine have heat sinks. They can handle high loads for short periods. They cannot handle high loads for long periods without derating. If your load is 50% for 1 hour, then 0% for 1 hour, the average load is 25%. But the peak is 50%. The generator must handle the 50% peak without overheating. If the cycle is 10% on, 90% off, the average is low, but the starting demand is still high. You must size for the peak, not the average. The duty cycle also affects fuel tank sizing. A generator running 2 hours a day needs a smaller tank than one running 24 hours a day.

Step 6: Apply Derating and Ambient Correction

Check the generator manufacturer’s derating curve. Most units lose 1 to 2% capacity per 5°C above 35°C ambient. For a 500 kVA generator in a 45°C environment, derate by roughly 4%.

Reason: Ignoring derating leads to thermal overload. The nameplate rating assumes a specific ambient condition. Real-world sites vary.

Look at the site. Is the generator in an open shed or an enclosed plant room? Is it in the sun or in the shade? If it is in a closed room with a heat load from the engines themselves, the ambient temperature inside that room will be higher than the outdoor temperature. You must measure the actual temperature inside the enclosure. If the outdoor temperature is 40°C, the enclosure might be 45°C or 50°C. Use the enclosure temperature. If the generator is rated for 35°C ambient and you are at 45°C, you lose capacity. A 500 kVA generator might only be safe at 475 kVA. If you buy a 500 kVA unit and your load is 490 kVA, you will overheat and trip.

Step 7: Select the Final Rating

Add the derated continuous load and the peak starting demand. Compare against standard generator ratings: 25 kVA, 50 kVA, 100 kVA, 150 kVA, 250 kVA, 500 kVA, 1000 kVA.

Choose the next standard size above your calculated peak. If your peak is 520 kVA, select 600 kVA. Do not round down.

Reason: Standard sizes match supply chain availability and spare part compatibility. Custom ratings increase cost and lead time.

If your calculation gives 520 kVA, you need 600 kVA. If your calculation gives 501 kVA, you need 600 kVA. There is no 505 kVA standard rating. You must go to the next size up. This gives you a safety margin. It also allows for future expansion. If you add a new machine next year, you might need 550 kVA. A 600 kVA generator handles that. A 500 kVA generator does not.

Common Mistakes in Generator Load Calculation

1. Ignoring starting demand. Treating all loads as continuous underestimates peak current by 200% or more.

2. Summing all motor starts. Adding every motor in the plant to the starting load creates an artificial worst case that no single generator can meet.

3. Assuming 100% power factor. Motors draw reactive power. A 50 kW motor at 0.85 PF draws 58.8 kVA. Always convert kW to kVA using PF.

4. Neglecting ambient temperature. A generator sized for 25°C will overheat in a 45°C plant room.

5. Skipping the duty cycle analysis. A generator sized for 1-hour daily use will fail if the load actually runs 20 hours daily.

6. Relying on total connected load. This is the most frequent error. It ignores diversity and simultaneity.

Final Verification Step

Before issuing a purchase order, perform this check:

1. Print the calculation sheet with all inputs and assumptions.
2. Verify every kW value against the original nameplate.
3. Confirm the diversity factor is supported by operational data or site history.
4. Check the derating calculation against the manufacturer’s published curve.
5. Have a second engineer review the simultaneous start assumption.
6. Confirm the selected rating is available from at least two vendors.

Reason: A single calculation error can void the entire project. Verification catches arithmetic slips and logic gaps.

Documentation for Procurement

The calculation sheet becomes part of the procurement package. Vendors use it to confirm suitability, size the fuel tank, and specify the control system. Keep the original with the as-built drawings. If the facility changes, recalculate. A generator sized for 2023 production may be undersized for 2026 expansion.