GMCC KSN98V11VEZ3 Rotary Compressor: 9.8 cm3 Printed in Three Tables, and the Under-Developing Flag

By the Qishanr technical desk. Reviewed 2 November 2026.

Key facts at a glance

The GMCC KSN98V11VEZ3 is a single-phase 50 Hz rotary compressor for R32, printed in the 2024 edition of the manufacturer's rotary compressor catalogue at 12,000 Btu/h class capacity with a coefficient of performance of 4.18. Everything below is taken from that catalogue, and every figure in this article can be traced back to one row of it.

That last line is the reason this article exists. The row is complete and internally consistent, and the catalogue simultaneously marks it as under developing, which the footnote on the same page spells out as a remark legend. A buyer reading only the numbers will conclude the part is available. It is not, at least not on the evidence of this edition.

What does the model string fix, and what does it leave open?

GMCC prints a naming-rules page at the front of the catalogue, in two panels. The panel that governs the SN family reads: refrigerant, application field, series, displacement, power, motor specification, structure change, special specification. Read against KSN98V11VEZ3, the string resolves as follows.

Two things follow. First, the string fixes the refrigerant, the series, the supply and the displacement to the tenth. Second, it leaves the letters after the power code undefined on the naming page, and those are exactly the letters that separate the released rows from the flagged ones in the table below. A buyer who reads the string as a specification will over-read it.

The displacement field is worth stating plainly because it is the most common misreading: the catalogue prints the value as cm3/rev times ten, so 98 is 9.8 and not 98. The same rule turns 125 into 12.5 elsewhere on the same page.

Why does one displacement appear three times in one catalogue?

This is the central reading hazard for this model, and it is larger than the model itself. The 9.8 displacement is printed in three separate tables of the 2024 edition, under three different refrigerants or drives, at three different test conditions, with three different capacities.

The three rows are one displacement wide and a factor of two apart in capacity. Nothing about the displacement explains that. The refrigerant does part of the work: R290 carries roughly half the volumetric capacity of R32 at the same swept volume, which is why the 1480 W row is not a weak compressor but a different gas. The drive and the test condition do the rest: the middle row is a DC inverter machine rated on a seasonal basis, the last row is a fixed-speed machine rated at a single point.

Each row closes arithmetically on its own terms. 1480 divided by 475 is 3.116 and the table prints 3.12. 3070 divided by 784 is 3.916 and the table prints 3.92. 3020 divided by 722 is 4.183 and the table prints 4.18. The capacity in Btu/h closes too at the constant of 3.412: 1480 gives 5050, 3070 gives 10475 and 3020 gives 10304, and all three are printed exactly.

What does not close is any comparison across the three columns. A 4.18 next to a 3.92 invites the conclusion that the fixed-speed R32 row is the efficient one, and that conclusion is not supported, because a GX single point and a SEER 60 seasonal rating are not the same measurement. The article states the three numbers side by side precisely so that a reader cannot pick one up without seeing the other two.

What does the GX row print, and what does it not?

The row itself, transcribed from page 13 of the 2024 edition, reads:

KSN98V11VEZ3 9.8 3020 10304 722 4.18 25/400 280 8.1 9.8

Read in the order of the column headers, that is the typical model, the displacement in cm3/rev, the cooling capacity in watts, the same capacity in Btu/h, the power input in watts, the coefficient of performance as a ratio, the run capacitor as microfarads over volts, the overall height in millimetres, the suction line in millimetres and the discharge line in millimetres.

Eight fields that a selection decision normally needs are absent from the table entirely, and no amount of arithmetic recovers them.

The discharge line is worth a second look, because it is the one size in the row that separates platforms rather than capacities. Every row in the SN block up to 9.7 cm3 prints a 9.8 mm discharge line, and the flagged rows at the top of the block print the same 9.8. In the neighbouring D-series inverter table, the two rows that print a 12.9 mm discharge are also the two rows with the 260 mm height, so size and structure move together there. For this model the 9.8 mm discharge places it in the small-bore platform, and the specification document is the only place to confirm the joint type that goes with it.

Where does this row sit in the SN ladder?

The SN block on page 13 runs from 4.0 to 9.8 cm3 across eight rows, all on the same R32 heading, the same 220/240 V single-phase supply and the same GX test condition. That makes the block a clean ladder: one variable, displacement, and everything else held constant by the heading.

All eight rows close on capacity over input within one rounding step, and all eight close on the Btu/h constant. The ladder is not a smooth line, though, and the interesting part is where it breaks.

Capacity per cm3 of displacement runs 287.5, 301.9, 297.1, 301.3, 296.5, 300.6, 302.2 and 308.2 across the eight rows. The last value is the highest in the block, and it is 3.9 per cent above the nearest released row below it. That is a small difference in absolute terms and a real one in direction, and it belongs in the flagged group rather than in the released group, which is the subject of the next two sections.

Why do the last three rows step up together?

The bottom five rows and the top three rows of the block are not separated by displacement. They are separated by three fields that move as a set.

Three things changed at once: the run capacitor voltage rose from 370 to 400, the height rose from 250 to 280, and the coefficient of performance rose by between 0.16 and 0.30. Because all three move together across exactly the same boundary, the table cannot tell a buyer which one causes the efficiency gain. It may be the motor, it may be the larger shell, and it may be a different valve arrangement inside the same bore. What the table does support is a platform statement: the SN family changes platform at 8.9 cm3, and everything above that point shares a 25/400 capacitor and a 280 mm shell.

The economics of the step are worth putting in numbers, because they are the closest thing to a selection advantage this row offers. Moving from 8.6 to 9.8 cm3 buys 18.4 per cent more capacity for 14.0 per cent more displacement and 9.7 per cent more input power. Capacity per cm3 rises 3.9 per cent, and the coefficient of performance rises from 3.88 to 4.18. For a condenser unit that is short of capacity at the top of its modulation range, that is the correct direction. For a unit that spends most of its life at half load, the single GX point says nothing about whether the gain survives.

What does the under-developing flag mean for a buyer?

The remark column on page 13 carries a triangle against KSN89V11VEZ3, KSN92V11VEZ3 and KSN98V11VEZ3, and the footnote at the foot of the same page defines it as under developing. The same remark and the same footnote appear on page 11 for the inverter rows and on page 9 for the R290 block, so this is a catalogue-wide convention rather than a formatting accident.

The three flagged rows are exactly the three rows of the higher platform. That is the single most useful sentence in this article: the efficiency step described above exists in the catalogue only on builds that the catalogue itself marks as not yet released. The released part of the SN block ends at 8.6 cm3 on a 3.88 coefficient of performance and a 250 mm shell.

For a buyer this changes the question from which one is better to whether this one can be supplied. The honest answer from the catalogue alone is that the design exists, the rating exists, and the release does not. A purchase enquiry should therefore ask for the release status and the lead time in the same message, and should carry the 8.6 row as the fallback in case the answer is that the 9.8 is not yet orderable. Nothing in this article should be read as a claim that the part is in production.

How does it compare with the nearest released row?

The comparison that matters operationally is between this row and the nearest row below it that carries no flag.

Two of those lines are the ones a system designer will act on. The shell grows by 30 mm, which is a mounting and enclosure question rather than a thermodynamic one, and the discharge line does not change, which means the piping design transfers without modification. The third line that matters is the last one: the better row is not the available row.

There is a second comparison worth making, against the inverter row of the same displacement. The DC inverter builds at 9.8 print 3070 W at 784 W in the base case, which is 1.7 per cent more capacity on 8.6 per cent more input than this fixed-speed row, at a SEER 60 condition rather than at GX. The capacity being almost identical while the input differs by nearly a tenth is the clearest illustration available that the printed coefficient of performance is a property of the measurement basis as much as of the machine. A buyer who needs the inverter behaviour should select on the inverter row and accept its own condition; a buyer who needs a single-point rating should stay in the GX block. Neither decision is served by comparing the two coefficients.

What do the supplier pages print, and where do those numbers come from?

The supplier site carries pages for several strings in this family, and they do not all agree with the catalogue. Reading them is still worthwhile, because they show which strings the trade is actually being offered, but each one has to be checked against the row rather than trusted.

KSN92V11VEZ3 is the one that agrees. Its page carries a summary that matches the catalogue row on all four numbers, and it describes the machine as fixed speed, which is consistent with a GX single-point rating. It is the only one of the five that a buyer can use as a cross-check.

KSN89D21UER31 is the most instructive failure of the five, because the page does not merely print numbers that differ from the edition. It attributes them to it. The summary says the manufacturer's rotary compressor manual lists 2750 W at 720 W for a coefficient of performance of 3.82, and the 2024 edition carries no row for that string at all: its only 8.9 cm3 row is KSN89V11VEZ3 at 2675 W, 655 W and 4.08. Either the page is citing a later edition, or it is citing a build that the edition does not print, and the page does not say which. A citation of a manual that the reader cannot find in that manual is worse than no citation, because it looks checkable.

KSN98D22UFZ is a different string and a different build, and its page prints 9.7 cm3 against a model code whose last-two-digit displacement field reads 98. The catalogue prints 9.8 for every row in that group. The page's other three figures close internally, so this one field is the outlier rather than the whole page being wrong.

The two 108-series pages are the clearest example of why the page is not the manual. KSN108D22UFZ prints 10.8 and 3455 W, while the 2024 edition prints two rows in that group, at 3410 W and 3380 W, on 865 W and 840 W. None of the page's four figures appears in the edition. That does not make the page wrong, since GMCC may have revised a build after the edition was printed. It does mean the page cannot be used as evidence for the row, and it should not be quoted as if it were.

One further defect runs across these pages and is worth naming because it misleads on the supply question: the field labelled Voltage carries the string DC INVERTER on the inverter pages rather than a voltage. That is a drive description sitting in a voltage field, and a buyer reading it for a supply will get no supply. The voltage has to come from the catalogue heading or from the specification document.

What duty does this row belong to?

The application field of the string is blank, and the naming page defines a blank field as compressors for air conditioning, with R for heat pumps, L for freezing and cold storage, E for heating and G for cabinet air conditioning as the alternatives. Four independent facts on the page agree.

The refrigerant is R32, which the catalogue uses for the air-conditioning and heat-pump blocks. The test condition is GX, which is a cooling rating and not a freezing one. The block sits under a page whose Chinese header names air conditioning explicitly. And there is no L-prefixed row anywhere in the edition: a scan of the whole catalogue for the freezing and cold-storage prefix returns nothing at all, which means the family convention is not being applied to this model in some hidden way.

The label refrigeration compressor therefore does not hold for KSN98V11VEZ3 in the sense the manufacturer uses. Refrigeration in English covers any vapour-compression duty, including air conditioning, so the phrase is not false in ordinary speech. It is false as a catalogue classification, and a buyer who orders on the phrase rather than on the string may receive a machine rated for a duty envelope they did not intend. The correct description is a fixed-speed R32 rotary compressor for air conditioning, rated at the GX condition, 3020 W at 4.18.

The application string on the supplier pages needs the same caution. It reads AC, drier, refrigeration, chiller, etc. on every product page in the catalogue of the site, including pages for models whose rows are plainly air-conditioning rows. It is a template line rather than a model-level statement, and this article does not use it as evidence for anything.

How should a buyer specify this model?

Six checks, in order, each of which has bitten a buyer in this family.

The order matters. Checking the supply before the release status, for instance, produces a confirmed voltage for a part that may not be orderable. Checking the release status first turns the remaining five checks into work that will actually be used.

For a tender, the useful phrasing is to require the manufacturer's release letter and a rated-point certificate naming the test condition, because neither is in the catalogue and both change the answer. A quotation that cites capacity without naming GX cannot be compared with a quotation that cites a seasonal rating, and the difference between the two is larger than the difference between any two rows in this article.

Where can this row be checked?

The row is on page 13 of the 2024 edition of the manufacturer's rotary compressor catalogue. The same edition prints the naming page at the front, the R290 block on page 9 and the inverter block on page 11, which are the three tables this article compares. The catalogue is distributed through the manufacturer's product centre and through the downloads page on the manufacturer site, and a distributor-hosted copy of the 2024 manual is also available for readers who want a second host for the same file.

For readers who want the family context rather than the single row, the site's GMCC rotary compressor range page and the category overview place this model inside the wider rotary compressor category, and the same displacement read at a different letter is worked through in the R410A example on the blog side. A 9.8 cm3 fixed-speed build is also the size at which buyers start weighing rotary against scroll, and the GMCC scroll line is where that comparison begins; it is a different architecture with its own catalogue, so the numbers do not transfer. None of those pages carries the rating, and none of them should be used instead of the row.

For the inverter alternative at the same displacement, the supplier's own page for KSN98D22UFZ is the closest thing the trade currently has to a product page for a 9.8 build, and its displacement field disagrees with the catalogue, so it is a starting point rather than a reference. The sister page for KSN89D66UKRC3 is in the same position.

Which standards and approvals frame the purchase?

A rotary compressor of this class is certified as a motor-compressor, and the safety standard that a test house cites for that component is IEC 60335-2-34, which is the part of the household appliance series written for motor-compressors. Naming the standard is not a certification statement: this article makes no claim that KSN98V11VEZ3 holds any certificate, because the catalogue prints no approval marks and the row cannot support one.

On the refrigerant side, any machine using R32 in the European market is placed on that market under Regulation (EU) 2024/573, which is the framework that governs fluorinated gases and the equipment that contains them. The regulation is where a buyer's compliance team will look, and it is cited here as the framework rather than as a statement about this model's charge or its classification.

The demand context that makes an air-conditioning compressor of this size worth selecting carefully is set out in the IEA's The Future of Cooling, which is the reference most often reached for when a specification needs to explain why part-load behaviour matters. This article takes the direction of that argument and none of its numbers.

Frequently asked questions

Is KSN98V11VEZ3 a refrigeration compressor?

Not in the manufacturer's classification. The application field of the string is blank, which the naming page defines as air conditioning, the refrigerant is R32, the rating is at the GX cooling condition, and the catalogue contains no freezing and cold-storage row at all. In ordinary English the phrase is broad enough to include air conditioning, which is why the label circulates. As a catalogue class it is wrong, and the row is the authority.

Is the 4.18 coefficient of performance better than the 3.92 of the inverter rows?

They are not comparable. 4.18 belongs to a fixed-speed row rated at a GX single point on page 13; 3.92 belongs to a DC inverter row rated on a seasonal basis on page 11. The capacity of the two rows is within 1.7 per cent of each other while the input differs by 8.6 per cent, which shows that the difference is in the basis of the rating rather than in the machine. Compare rows inside one table and one condition, or ask for both to be re-rated on the same basis.

Can I order this model now?

The catalogue does not say so. The row carries the triangle that the same page's footnote defines as under developing, and the two rows below it carry the same mark. That is a statement about release, not about capability. Ask the manufacturer for the release status and the lead time before you design the unit around this rating, and keep the 8.6 cm3 row without a flag as the fallback. The news index on this site is where a release announcement would appear if the edition has since been superseded.

What does the trailing 3 in the string mean?

Nothing the catalogue defines. The naming page covers the refrigerant, the application field, the series, the displacement, the power, the motor specification and the structure-change positions, and it stops there. The characters after the power code are not decoded anywhere in the edition. Treat them as part of the part number and nothing more; do not read a generation or a revision out of them.

Does the 12.9 mm discharge line of some inverter rows apply here?

No. This row prints 9.8 mm, and so does every other row in the SN block up to 9.7 cm3. The 12.9 mm discharge appears on the two inverter rows with a 260 mm shell, which is a different structure. The discharge size is one of the few fields in the row that separates platforms rather than capacities, so it is worth copying into a piping drawing exactly as printed.

Why does the supplier page for KSN98D22UFZ say 9.7 cm3?

The page prints 9.7 while the model code carries 98 and the catalogue prints 9.8 for every row in that group. The page's capacity, input and coefficient of performance close on each other, so the outlier is the displacement field rather than the page. Use the catalogue figure for the group and ask the supplier to correct the page; do not average the two.

Which build is the nearest replacement if this one is not released?

The 8.6 cm3 row, which carries no remark. It prints 2550 W at 658 W for a coefficient of performance of 3.88 on the same R32 heading, the same 220/240 V supply, the same GX condition and the same 9.8 mm discharge line. It is 18.4 per cent down on capacity and 30 mm shorter in the shell, and it is available on the evidence of the catalogue.

What does the 25/400 capacitor figure mean in practice?

It is the run capacitor, printed as microfarads over volts, so 25 uF at 400 V. The released rows below 8.9 cm3 print 25/370 or smaller, and the flagged rows print 25/400. A replacement capacitor must match both numbers, and the voltage rating is the one that is easy to under-specify when a service engineer substitutes from stock.

What supply does this compressor need?

Single phase, 50 Hz, 220/240 V, which is the heading of the column the row sits under. The string's power position carries V, which the naming page defines exactly that way. The same series also appears on 100 V and 230 V builds, so the supply cannot be inferred from the series or the displacement, only from the power position of the string or from the column heading.

Sources