Define one motor-start event, calculate screening kVA from declared current, and request configuration-specific voltage, frequency, and recovery evidence.
A motor kW or hp value is not enough to size a set for starting. Define one start event, calculate screening kVA from the stated start current, and require voltage and frequency traces for the exact quoted set; stop the comparison when those bases differ. This RFQ method does not select the final set, set a universal dip limit, or design motor controls, cables, switchgear, or the grounding system.
Define the start event before choosing a generator
The largest motor may not cause the hardest start. A smaller motor can be worse if its load torque is high or it takes a long time to reach speed. Loads may also be online before it starts. Some of those loads may trip during the dip.
Give each event a fixed ID, such as MS-01. State what is running just before the start. One full event joins five records:
- the exact motor and driven machine.
- the starter method, settings, bypass logic, and start sequence.
- the loads already connected to the generator.
- the exact set build and its site rating basis.
- the test point and pass limits for voltage and frequency.
List two events when two start paths are real. One may be a normal start. The other may be a restart while key loads stay online. Do not merge them into one vague “largest motor” case. The industrial generator sizing guide shows how this event fits into the full load list.
Build the motor-start input sheet
Gather stated or measured start data before you use any current factor. Mark a missing field OPEN on the work sheet. Do not hide a guess in the quote.
| Field. | Controlled input to send. | Why it changes the review. |
|---|---|---|
| Motor ID. | Maker, type, model or plate ID, rated output, rated voltage, frequency, phase, rated current, power factor, and efficiency if known. | Keeps the current, power, and system basis tied to one motor. |
| Start-current proof. | Locked-rotor current, the maker's start curve, a starter study, or a current trace from a test. Add the source and file version. | Shows the size and length of the current rise without one fixed factor. |
| Driven load. | Machine type, loaded or no-load start, torque-speed curve, inertia or other run-up data, and time to reach speed. | Shows whether the motor can bring the load up to speed. |
| Starter and controls. | Direct-on-line, star-delta, autotransformer, soft starter, VFD, or other method. Add the settings, switch point, bypass, current limit, and ramp where needed. | A starter name alone does not define the input load. |
| State before start. | Loads online, start order, auto restarts, key controls, and any load step at the same time. | Shows what the set is already carrying. |
| Power basis. | Required voltage, frequency, phase, wire scheme, and test point. | Stops formulas and vendor replies from using two different systems. |
| Site and duty. | Set duty, height above sea level, air or inlet basis, room or canopy layout, and fuel state where needed. | Ties the start reply to the set as offered for the site. |
| Pass record. | Loads that must stay online, base voltage and frequency, test point, low limits, return band and time, and the source of each limit. | Makes the allowed dip a job input, not an online rule. |
Use the generator voltage and frequency checklist for the target market and power system. If that system is still open, the start result is not ready to compare.
Treat starting kVA as a screening quantity
Starting kVA turns stated voltage and start current into apparent power for the same event. It gives each vendor one common input. It does not pick the set.
For a balanced three-phase event:
Screening kVA = √3 × line-to-line voltage × starting line current ÷ 1000
For a single-phase event:
Screening kVA = voltage × starting current ÷ 1000
These equations only convert units. They do not show how long the current lasts. They also leave out start power factor, load torque, loads online, engine response, and the way the set controls voltage and speed. Those facts decide whether the motor reaches speed and other loads stay on.
Do not compare start kVA with an unlabeled “maximum,” “surge,” or book value. First match the rating name, time basis, duty, voltage, frequency, phase, and site. The rated-versus-peak generator power guide shows how to do that.
Use the BEAR generator catalogue to shortlist configurations with the required phase and rating basis, then require exact-set motor-start traces before treating any model as comparable.
Starter type changes the event, not the evidence burden
A reduced-voltage starter, soft starter, or VFD can change the current seen by the set. The starter type does not tell you how much. It also does not prove that the motor can bring the load up to speed.
| Starting method. | Event data that must be returned. | Review boundary. |
|---|---|---|
| Direct-on-line. | Start current over time, start torque, run-up basis, and start order. | Do not swap the real record for a fixed full-load-current factor. |
| Star-delta or other switch starter. | Switch method and time, open or closed switch, current and torque in each step, and bypass state. | The switch may cause a second dip. Ask for the full event. |
| Autotransformer or reactor. | Tap or setting, switch logic, voltage at the motor, input current, torque, and time to reach speed. | Less motor voltage can also mean less torque. |
| Soft starter. | Current limit, ramp, kick start or boost, bypass logic, load state, run-up trace, and current at the supply side. | The set point and driven load shape the event. |
| VFD. | Drive model and rating, front-end type, ramp, current limit, input current and harmonic data, bypass mode, and the maker's guide for set power. | Current at the motor side is not always the load seen by the set. |
If a vendor changes a starter setting to make the set pass, treat that as a new setup. Check that the motor, driven machine, process, starter maker, and system designer allow it. Only then review the new result.
Separate initial dip, minimum voltage, and recovery
“Voltage dip” can mean more than one thing. Ask for voltage and frequency over time at a named test point. Record the level before start, the first move, the low point, the return band and time, and the stable state after start.
Ask the vendor to fill these fields for each event:
- pre-start generator load, voltage, and frequency.
- test point and the way the trace or report was made.
- voltage at the start, its low point, and the time to that point.
- low or high frequency as needed, plus the way it returns.
- the set return band, time to enter it again, and proof that the set then stays stable.
- loads that stayed on, trips or control actions, inputs, and items left out.
There is no one safe dip or return time for every motor and load. A contactor, control supply, VFD, or process control may set the real limit. State where each limit came from: the job spec, the device maker, a system study, or the engineer in charge. Keep that source with the event sheet.
Demand proof for the exact proposed generator set
Start results belong to the full set and the case in which it runs. An alternator start curve can help. But it does not prove the joint response of the engine, alternator, AVR or excitation, speed control, canopy, fans, and site-rated set.
Use a Motor-Start Event & Supplier Return Sheet with these key fields:
| Return block. | Supplier response required. |
|---|---|
| Input ID. | Event ID, buyer sheet version, motor and load record, starter settings, loads online, power system, site, duty, and pass limits used. |
| Set offered. | Set model and build. Name the engine, alternator, AVR or excitation, speed control, voltage, frequency, phase, canopy and cooling layout, and key set loads. |
| Rating basis. | Rated kW and kVA terms, power factor, duty, output left at the site, and load on the set before the start. |
| Method ID. | Software run, sum, curve, test, or expert review. Give the file title and version, input file, issue date, reviewer, inputs, and items left out. |
| Start result. | Current or apparent-power trace used, run-up basis, starter action, and whether the motor brings its load up to speed. |
| Start trace. | Voltage and frequency over time, test point, low values, return band and time, state after start, and pass basis for the stated limits. |
| Change control. | Limits of the set build, open points, allowed part swaps, changes that need a new review, and vendor sign-off date. |
Ask for the current motor-starting file that applies to the exact part or set in the quote. A generic maker guide does not prove the offered set or replace a job review.
Keep the proof with the quote version. The generator document checklist helps keep the sum, drawing, data sheet, and offer on the same set build.
Stop the comparison when evidence bases differ
A larger headline kVA does not fix missing motor data or mismatched start proof. Put the quote on hold when any of these gaps remains:
- the motor or starter input is a generic multiplier rather than a declared or controlled event record.
- phase, voltage, frequency, duty, site condition, or loads already online differ between offers.
- one vendor gives an alternator curve while another gives a study or test for the full set.
- the engine, alternator, AVR or excitation, speed control, or rating basis is not named or can change with no new review.
- a steady rating is used as proof of the motor start.
- voltage dip, frequency trace, test point, return band, or return time is missing.
- the starter setting used in the analysis is not the approved setting for the motor and driven machine.
- inputs or items left out make the event hard to repeat.
Use COMPARABLE only when the event, set build, method, and reply share one fixed basis. Use CONDITIONAL when a named input is still open. Use NOT OFFERED when the vendor declines the event, and OPEN when proof is missing. These states control the quote; they are not design sign-off.
Skilled project staff should check the final set choice, power design, motor and load, controls, trips, and site plan. For a first BEAR set review, send Miya the filled start sheet, power basis, site data, start order, order size, and target market. Do not assume a model or start result is on offer until BEAR gives proof for that exact set build.
