Starship Volumetrics II:  ChronoVolumetrics

Starship Volumetrics I
Starship Volumetrics II:  Chrono-Volumetrics
Starship Volumetrics III:  Crew Densities

Recommended Reading:
Volumetrics 101 by 'BHMM'

Below, I show the volumetrics chart with a roughly-chronological assortment of ship classes.  Class chronology discussions could constitute an entire website's worth of material (there've been heated debates on the topic of what particular classes came before others), but suffice it to say that my list will primarily be ordered based on the roughly-chronological nature of registries, as well as on the general appearance of the vessels.  Smallest-number canon-source registries are provided, and the Volumetrics table has been modified slightly to make room for that.  Asterisks indicate some issue with the data.   I'm also putting a rough estimate of when the class ship would've been built.   Relatively extensive notes follow.

Starship Registries:  Chronologies and Ramifications

Quick Reference:
I.  General Issues
II.  Known and Derived Quantities
   A.  The TOS/TMP Era
   B.  The Starships of 2161
        1.  The NCC-0 Assumption
        2.  The NCC-200 Estimate
        3.  Ramifications of NCC-200 and 2000
   C. The TNG Era
        1.  The Galaxy Class Starships
        2.  The Danube, Defiant, and Intrepid 
        3.  NCC Assignment
        4.  Shipbuilding and Fleet Size
III.  Class-Specific Comments

I.  General Issues

There are a number of questions about the finer points of starship registries.  Potential confusion about how they work can arise along numerous lines.   Are they assigned chronologically, or in batches per class?   Is there any skipping of numbers?   Are they assigned when a ship is planned, or when it's commissioned?   And who really cares?  ;)

Some questions do have answers. What we do know is that as time passes, registries become larger numbers.  And, while registries can be reused (as in the case of 1701 with letter suffixes, and 74205 without them), this is exceedingly rare.  Further, given that the Yamato was identified as both 1305-E and 71807, we can conclude that such ceremonial registries have a real spot in the registry list . . . in other words, 1701-D would be the alternate designation for 71701 or something like that.

Also, there seem to be few (if any) cases where we see that many ships of a certain class all have registries within a certain block of numbers.   The Galaxy Class, for example, shows registries ranging from 70637 to 71832.  So, unless we conclude that the block of numbers has huge gaps, or that there are around 1200 Galaxy Class ships in service, it would seem that the ships are numbered based on chronology alone.  More on this matter appears below.

In any case, with the chronology generality established, we can get into specifics.

II.  Known and Derived Quantities

A.  The TOS/TMP Era:  Roddenberry's Enterprise, Et Cetera

Any discussion of registries should start with NCC-1701 . . . the USS Enterprise.  Let's ponder a couple of facts:

1.  Roddenberry wanted the Enterprise to be a ship with "some history", and her first launch is generally established as 2245, some twenty years prior to TOS.  

2.  In "The Cage"[TOS], set in 2254, Captain Pike and his crew encounter what they believe to be survivors of a ship which crashed 18 years prior (2236).   One of the Enterprise crew explains that the survivors won't believe how fast they'll get home, because "our new ships" are faster.   Since the survivors were loading up to board the Enterprise, it follows that the Enterprise was one such new ship.

The above would give us a likely build date for the Enterprise sometime in the 2236-2254 range, corresponding nicely with the official Roddenberry 2245 date which is smack in the middle of the range.  ( Of course there was Admiral Morrow's line in Star Trek III that the Enterprise was 20 years old . . . but that line contradicts all others.   It had been almost that long since the TMP refit, which occurred at least a decade and a half after "The Cage".)

Now, let's factor in a couple of other points:

3.  In "Court Martial"[TOS1], set in 2266, the ship list behind Commodore Stone shows NCCs well into the 1800s, skirting 1900:

4.  Last but not least, there's Star Trek:TMP (2272), which features comm chatter including a reference to the USS Entente, NCC-2120.   (Incidentally, this ship name and registry come from the Franz Joseph Tech Manual).

The first two of the above data points lead us to a logical build date for the Enterprise (no bloody A, B, C, D, E, or NX).   The latter two points give us some temporal context for the E's registry.

The data points above are in good agreement with one another.  For instance, if we take the registry of 1701 and presume it was assigned in 2245, then that's about 84 years after the founding of the Federation, established in TNG as occurring in 2161.  Half of that year-count is 42 . . . half of that registry number is 850ish.  The halving maneuver suggests that, on a purely linear scale (which is unrealistic, but useful as a very rough ballpark), then a vessel with a registry of  NCC-850 being built circa 2203 (plus or minus some years depending on the actual rate of ship construction).

If we check further, we find that this relation holds.  Assuming Commodore Stone's listings were the most recently built ships of the fleet, we have NCC-almost-1900 in existence in 2266.   Halving again, we have 950 ships and 52 years . . . making NCC-950 appear in 2213.  The TMP Entente reference is from 111 years after the founding of the Federation, giving us a halving of 55.5 years and 1060 ships.  That would put NCC-1060 at about 2216, which also roughly fits the above data points.

Interestingly, these points put the USS Constellation in an interesting light.   The Constellation, seen in "The Doomsday Machine"[TOS2], was a Constitution Class starship . . . her registry was NCC-1017.   Barring massive reconstruction of an earlier ship or a registry rarity such as was seen with the two Defiants of DS9, uch a low registry would imply that the Constitution Class was actually quite old . . . appearing sometime in the 2210 range.

This concept actually makes a lot of sense, though.   Unlike Miranda Class contemporaries, the Constitutions were a seemingly retired class by the 24th Century . . . even the Enterprise-A, a new-build Constitution of the 2280s, was retired in the 2290s.  If the class itself was 80 years old by that point, it would make sense to retire it, and might even explain the trouble with its construction seen in Star Trek V.

Some might disagree, citing Larry Marvick from "Is There In Truth No Beauty?"[TOS3].  This fellow, who could not have been more than fifty years of age, was referred to as "one of the designers of the Enterprise".  Some conclude that this makes him one of the designers of the Constitution Class, and others mislabel his work as being on the ship's engines.  These ideas would limit the age of the Constitution Class significantly, to the point that we couldn't expect a Constitution Class ship older than 2245.  However, neither view is accurate . . . Marvick's role with the Medusans in the episode is specifically labelled as regarding instrumentation (presumably related to the Medusan specialty of navigation), and Scotty bets him a bottle of scotch that he can't handle the Engineering controls he designed.   Thus, Marvick does not limit the age of the Constitution Class . . . just the control hardware in use in the 2260's.   Similarly, the computers of the Enterprise in the 2260's were based on Daystrom's duotronic computer design, developed around 25 years prior . . . but that doesn't alter the possible age of the class, either.  Indeed, it's possible that Marvick designed interfaces that would work best with the new duotronic computers, and that these were both incorporated into the ship during a refit (presumably prior to 2254's "The Cage", which showed interfaces quite similar to those of the TOS era).

Thus, there is no discrepancy, and the three  independent confirmations based on registry both confirm Gene Roddenberry's edict about the Enterprise's age.  Based on her registry, she would have been almost ten years old when we first saw her in "The Cage", set in 2254.   

However, while the above data points are all in agreement relative to each other, the dating scheme I used to get the dates is just an assumption, and here's why:

B.  The Starships of 2161

1.  The NCC-0 Assumption and the Age of the Constitution

The above is based on a peculiar assumption . . .  that "NCC-0" would correspond to a date of 2161.   This is certainly not the case.   It is likely that ships such as the NX-Class will be absorbed into the early Federation fleet providing an early boost to the NCC numbers.    That seems especially likely given that NX-02, sister ship to Enterprise NX-01, was seen about half-built at the end of Enterprise's second season, and it was said she wouldn't be finished for another 14 months.  Given that there were also other classes of warp-driven Earth ship, not to mention ships of the Vulcan, Tellarite, and Andorian fleets, this does imply a somewhat low rate of construction by the time of the Federation's founding.

Also consider that the USS Essex, NCC-173, was lost in 2167, just six years after the founding of the Federation.  To have an NCC that high, the Federation would've had to have been building almost 30 ships per year.   This may not sound like much, but if NCC-1701 only appeared in 2245, then using the NCC-0 assumption the ship construction rate from 2161 to 2245 would've been only 20 ships per year, and this during a time of cold war and expansion.  Further, that completely ignores the fact that the Daedalus Class ship was lost in deep space, which it would've taken her a long time to reach.

So why use the NCC-0 Assumption as a default?  Simple:  we cannot know precisely how many ships will be absorbed into the early Federation Starfleet.   There could be a few, or there could be hundreds.   It could just be Earth ships, or those of Vulcan, Andor, and Tellar as well . . . and we don't know what types are absorbed.  

The chronological implications for pre-TOS-era starships can be profound, since the known values of the Enterprise, Entente, and the "Court Martial" chart can be used to derive averages.   Focusing on the Entente, for instance, the average rate of ship construction from 2161 to 2272 would be about 19.1 ships per year, using the NCC-0 assumption.  However, even if we concluded that the Federation Starfleet had absorbed, say, 500 ships when it started up (as an 'NCC-500 assumption'), then this average rate would become reduced.  From NCC-500 in 2161 to NCC-2120 in 2272 would give us 1620 ships in 111 years, for an average rate of 14.6 ships per year.

As a result, if we wanted to learn when the Constellation (NCC-1017) was built we would find a large spread of possible values, as seen to the left.  (We already get a significant spread just from the NCC-0 assumption . . . we can calculate it based on the average as counted from 2161, or backward from the Enterprise build date, or backward from TMP . . . each will give a slightly different value.  To the left, the lines are drawn from NCC-0 or NCC-500 to the Entente from TMP, with various other ships (including the Enterprise) and other years also marked.)

Using the NCC-0 assumption, then we could simply use the average in various ways from any of the available data points:

1.  If we started at 2161 and used the average rate from the Entente, we would arrive at 2214.

2.  Or, if we worked backward from 2245's NCC-1701, we would arrive at 2210. 

 The NCC-500 idea allows a similar spread, but it falls within a different decade altogether.

1. At 14.6 ships per year, the Constellation would've been built circa 2196 as counted from 2161.

2.  Or, if we worked backward from 2245's NCC-1701, we would arrive at 2199.  

(Incidentally, the Daedalus Class, like the Essex mentioned earlier, was retired in 2196.  If the NCC-500 dating scheme were correct, it could imply that there was a direct line between the Daedalus and Constitution.   Further, if we were to use the USS Eagle (NCC-956), a Constitution Class ship included (but not visible) on a chart from Star Trek VI, then the Constitution Class would be pushed back to 2192, making her a brief contemporary of the Daedalus Class.) 

In any case, we know that from NCC-1701 in 2245 to NCC-2120 in 2272, at least 419 ships were built.   This gives us a rate of just 15.5 ships per year in that era, minimum.  If we had another ship NCC and date between 2161 and 2245 to compare it with, we could get a pre-2245 rate, but we don't have that.   However, from 2161 to 2245, there would've been at least 1701 registries per those 84 years, which would be 20.25 ships per year using NCC-0.  Using NCC-500, that's 14.3 ships per year, which is more consistent.

1a.  No NX-01 Assumption?

Since we know that the Enterprise NX-01 was launched in April, 2151, one might ask why I do not gauge NCC numbers as starting from there . . . i.e. taking NX-01 of 2151 and NCC-1701 of 2245, and making estimates based on the 1700 numbers and 94 years.  There are several interrelated reasons why this ought not be done.   First, the NX-01 registry is based on the fact that the vessel is of the NX Class.  The Columbia is NX-02.   If we concluded that the Earth fleet starship count in the early 2150's was only two based on that, however, then we'd be wrong . . . there's been mention of entire other classes of warp-driven starship, not to mention non-NX vessels such as the Intrepid and Shenandoah.  (Somewhat analogous would be the numbering scheme of US Navy ships.)

Further, we know that the first Federation Starfleet registry of 01 ought to belong to a vessel named Dauntless.  Whether this was an NCC number or an NX number is unclear . . . the NX-01-A from "Hope and Fear"[VOY4] was an experimental vessel, and might've had an NX prefix despite being the 'dash-A' of an NCC-01.  In any case, while the NX-01-A was only an alien vessel disguised as a Federation ship, we have no reason to conclude that the Voyager crew would've been so ignorant of history as to forget the name of the original 01.

(Rumors that the original 01's real registry was actually CNH 320 are unfounded.)

The Dauntless also helps us eliminate a potential source of worry.  After all, if the Federation Starfleet had arbitrarily started with hull number 101 (a la NASA's Enterprise OV-101), then the counts would be off.  However, since there was obviously a hull number 01 in the Federation Starfleet, then we may assume that they did not then skip to a three-digit number.

2.  The NCC-200 Estimate

So, is there any way we can even guess at how many ships were part of the Federation Starfleet in 2161?  Unfortunately, not really.  We can presume that the Essex 173 was at least a few hundred light-years out from Earth, given that she was destroyed somewhere fairly close to where, in 2368, the USS Denver struck a mine left over from the Cardassian conflict.  (Chakotay's home planet on the Federation-Cardassian border was canonically stated to be thousands of light-years from Earth in "Pathways", though maps from DS9 indicate that the border could've come far closer to Earth in some places.)  If so, then the Essex had probably been in space for at least a couple of years as a lower limit.   Thus, the very latest date for the construction of the Essex would've been about 2165.  And, assuming a rate of starship construction no greater than that of the later Federation (a tricky assumption given real Earth history, but no matter), then the largest lower-limit NCC for an absorbed ship would've been about 110.

As for the largest upper limit number, there's really no clear cap.  I'd personally be averse to having the Constellation or any other Constitution Class ship in existence circa 2161, but given that the basic design isn't all that different from the NX Class it could theoretically be possible.  But, that would require a thousand ships between the Vulcans, Andorians, Tellarites, and Terrans . . . a figure which seems awfully high, especially considering the small number of Earth Starfleet ships just ten years earlier.  Each of the other races would have to have at least 300 ships each.  This seems wildly implausible given certain events from 2150's.   With the disbanding of the Vulcan High Command in 2154 ("United"[ENT4]), for instance, Vulcan's naval forces were significantly reduced.   Further, when a 128-ship fleet was required, the Vulcans could only dispatch 23, and the slow Terran vessels, if they were involved at all, were probably low in number as well.   Assuming a full 128 starships were collected together, then most would've had to have been Andorian and Tellarite.  But even the Andorian fleet is unlikely to possess 300 ships.

The Earth Starfleet is not at all likely to have more than 100 proper starships by then.  The number of Tellarite ships is unknown, but the Vulcans and Andorians could be assumed to have had a rough parity given their cold war, at least prior to 2154.   Just as a general impression, I doubt that there would be 500 starships among the lot of them.  Thus, I'd place NCC-500 as a rough upper limit.   After all, at a construction rate of 20 ships per year and assuming an average starship lifespan at the time of about 50 years, the fleet circa TOS/TMP would've consisted of about 1000 starships.

Another consideration would be ship lineage and fleet makeup.  Though we've been explicitly told that alien ships and forces are absorbed into later Starfleet when a planet's admitted ("Rapture"[DSN5]), we haven't seen much to indicate that.   With the exception of a few oddball vessels like the Norkova (and perhaps the Aurora), most Federation ships look basically like evolutions of the Earth ships of the 2150's.   We could therefore estimate that the Earth fleet would comprise a significant portion of the early Federation starships of the 2160's, with their lineage being carried on (with some alien technological boosts) for centuries thereafter.

Finally, there's the Daedalus Class herself.   Though the class has not appeared in Enterprise, we know several facts.  The Essex crew count was 229, and a retirement date of 2196 is given for the class.   The size is apparently smaller than Enterprise, and the ships have a hull configuration which is based on simpler shapes . . . cylinders and spheres.  However, the ships have a separate engineering hull like many later starships, and, like the TOS-era Constitutions, do not apparently feature glowy bits on the outer nacelles.  They also do not seem to have a navigational deflector dish.

With the NX Class, we find a spaceframe that can last 120 years ("E2"[ENT3]), a smaller crew, and a larger, more complex hull design.  Further, the more recent ships (NX and the ship type of which the Intrepid is a part) do feature a navigational deflector dish.

The way I see it, there are two possibilities.  Either the Daedalus Class is older than the NX Class, with a hull configuration demonstrating less-advanced shipbuilding techniques (which just happen to have some details carried on in some later ships), or else the Daedalus is a newer, simpler design with a compact planform intended to allow for a large number of ships to be constructed in a hurry . . . the sort of thing one might expect during, say, the Romulan War of the late 2150's.   Either notion fits in well with her crew count, retirement date, and hull configuration.

Further, either one would imply that the ships should've been in existence in 2161.   As a result, I would surmise that the Essex was probably around in 2161, which makes more chronological sense for that ship.   With her NCC of 173, that would imply that at least about 175 ships were absorbed into the early Federation Starfleet.  Preferring a nice round number and some maneuvering room beyond the Essex, I will estimate that some 200 vessels were registered with Starfleet in 2161.

3.  Ramifications of NCC-200 and NCC-2000

Given the above, we can conclude that from 2161 to 2245, 1501 starships were constructed or otherwise added to the Federation Starfleet, for an average of 17.9 ships per year.  From 2161 to 2272, that average drops to 17.3 ships per year.   Starting with NCC-200, that would place the approximate date of the Constellation as 2212-2214.  Or, working backward from NCC-1701 and NCC-2120, respectively, we'd end up with dates of 2206-2208.

Thus, we have some degree of certainty that the Constitution Class would've appeared circa 2210.

Now, note that if we projected the 20 ship per year count forward after 2245, we'd have 100 ships every five years, or 400 every 20 years.  That would put us at NCC-2101 by 2265, NCC-2501 by 2285.

However, 2285 saw us with NX-2000 just being readied for trial runs!  How could this be?  Well, we already know that there was a ship running around 13 years prior with a higher registry, so the problem is not with our estimates.  To explain NCC-2000 given the estimates and the Entente, we must find some reason why the ship would have such an anachronistic registry.   The solution is an obvious one . . . the ship was the keystone of the Transwarp Development Project.   We can readily allow for the idea that such a project, and the ship it centered around, were decades in development, and that the NCC number was held for the ship.  Given that the next known Excelsior has a registry in the 2500s, this seems to be the case.

But, if we keep projecting past 2300, then we find ourselves with NCC-2901 by 2305, and only NCC-4101 by 2365, which is the second season of TNG.  Of course, by that point, registries were in the 70000 range!  We also know that they could build no less than 40 ships in one year, as per Shelby in "Best of Both Worlds, Pt. II"[TNG4].  In short, after about 2300 or so, the TOS/TMP-era NCC projections don't work well at all, which leads us to:

C.  The TNG Era Registries

1.  The Galaxy Class Starships

The TNG Era, from our perspective, began with a starship that is completely useless for our purposes . . . NCC-1701-D, the ceremonial registry of the USS Enterprise.  The Enterprise-D, however, was a member of the Galaxy Class of starships, and we do have a few registries for them.

For instance, we have the class ship 70637, along with NCC numbers 71099, 71807, and 71832.  NCC-71807 was the Yamato, destroyed in 2365.  The Enterprise-D, judging by her plaque, was commissioned in late 2363, entering service in 2364.  So, we can assume that the class ship was commissioned no later than 2363.  The registries had reached 71807 by no later than 2365.

It would be nice to know the exact commission date of another Galaxy with a known registry, and we do, in a sense.  There are numerous indications that the Galaxy Class is brand-spanking new by the time of TNG.  For example, in "Contagion"[TNG2], Captain Varley of the USS Yamato pondered whether his ship's systems problems were possibly a design flaw or a result of insufficient testing.  This limits the ship to a reasonable timeframe prior to the Enterprise . . . that is to say, you don't start thinking of design flaws, even if you're explicitly "grasping at straws", when you start having trouble with your ship 5 or 10 years after its launch.  Further, we know that the prototype testbed vessel Pegasus, lost in 2358, had some systems aboard which were employed in the construction of the Galaxy Class.   More conclusively, the final design of the Galaxy Class dilithium crystal chamber is dated Stardate 40052 (at Outpost Seran-T-One, per "Booby Trap"[TNG3]).  Given the rough progression of stardates, this prevents us from concluding that the Galaxy was built prior to 2363.   Finally, the prototype design room at Utopia Planitia's Mars Station seen in "Booby Trap", dated Stardate 40174, shows a Galaxy Class ship under construction outside the window.   As the prototype for the class, this can only be the Galaxy herself.

Thus, we know with certainty that the Galaxy herself was constructed (or finished) and tested in 2363, with the Enterprise being built in the same year.  This allows us to narrow down the possible dates for the Yamato's construction . . . she could only have been built after the Galaxy in 2363, and after the Challenger (NCC-71099) as well.  But, she was built prior to her destruction (obviously) in 2365, and had probably been in space for at least a little while by that time.  A likely build date would be 2364, and the same for the Odyssey, whose registry is only 25 units distant.

2.  The Danube, Defiant, and Intrepid Classes

So, now that we've narrowed down the Galaxy chronology a little, let's take a moment to ponder other TNG-era examples.   For instance, we know that the Danube Class runabouts were commissioned in 2368, per dialogue in 2370's "Paradise"[DSN2] that they'd been commissioned two years beforehand.  And, we know that Voyager, commissioned in 2371, has an NCC of 74656.  The Sao Paulo, commissioned in 2375, had a registry of 75633.  From the Sao Paulo to Voyager is 977 units in four years, giving us a figure of 240 NCCs per year.  We could thus take the 200 NCC per year figure as a lower-limit ground.

But, if we extrapolate backwards, the 200 NCC-per-year figure would suggest a registry date for the Galaxy of 2351, and 2356 for the Yamato.  This would require that the class be some 12 years old by the time of TNG season one, which is quite contradictory.   The alternative is to compare Voyager's 74656 from 2371 with the Challenger's 71099 from 2363 or the Yamato's 71807, believed to be from 2364.  This would give us between 407 and 445 NCCs per year during the 2360s.  Such a figure is confirmed by the Rio Grande, the earliest known Danube . . . from the Sao Paulo to the Rio Grande gives us 454 NCCs per year . . . or a surprising 735 NCCs per year from the Rio Grande to Voyager.   Counting from the Yamato, any of those 400-range figures would put the Defiant as receiving her 74205 registry in 2369, which could be quite possible if she received her registry shortly before Sisko left for DS9 (his arrival at DS9 was on Stardate 46379.1, or almost mid-2369).   This would suggest that the Defiant Class Development Project took two years or so (starting in 2367 after the Borg incident), which is quite acceptable (and the maximum allowed by the canon anyway).  The 735 NCC per year figure, counting from the Yamato, would result in a Defiant NCC date of 2364 . . . which is obviously too early.

But why the difference?   After all, we're looking at over 400 NCCs per year in the 2360s, versus just 240 per year in the 2370s.  The answer is in DS9.  Romulan Senator Vreetak notes in "In the Pale Moonlight"[DS9-6] that the Federation's shipyards "are still being rebuilt", meaning some of them had evidently fallen victim to either the Klingons or the Dominion . . . more likely the former, given the long-standing treaty and sudden surprise conflict in the early 2370's.   Thus, we can presume that as of his comment in 2374, Federation shipbuilding had been slowed dramatically by damage to shipyards.

3.  NCC Assignment:  A Matter of Time

We can work on answering the question of when a ship gets her registry with the help of Voyager and Intrepid.  In Nemesis, the class ship of the Intrepid Class is given a registry of NCC-74600, just 56 units away from Voyager herself.  At a rate of 200 NCCs per year, this would mean that Voyager's registry was about three months distant from that of her class ship.  And, given the comment from "Relativity"[VOY5] that Voyager would be the first to test the warp core on deep space duty, we can assume that Voyager was the second ship of the class to be built.

So, if only three months passed between the class ship's NCC and that of Voyager, then we have some questions to ponder.  Is three months enough time to test a class ship, and then have another one built?   Possibly . . . after all, the much larger and more complex Galaxy Class ships could be built in less than a year, given the Challenger and Yamato.   Is that enough time to design a ship, build it, test it, and then build another?   Almost certainly not.  

In other words, the Intrepid-Voyager issue limits us to certain conclusions.  Whenever we conclude that the class ship got her registry, Voyager had to have gotten hers at a similar or earlier point in her lifetime, but not later.  For example, if we concluded that the Intrepid got her registry when the keel was laid, then we could assume Voyager got hers either when the decision to build her occurred, or else when her keel was laid.   However, it would be unlikely that three months passed between the Intrepid's keel-laying and the launch of Voyager.

Given the Excelsior, we could conclude that the NCC is given when a ship is planned for building.  However, the Defiant Class Development Project which began in 2367 does not seem to have resulted in an NCC until two years later, when the ship would've been built or completed.  Further, assuming that the standard starship development project time is at least two years, then the Intrepid's registry would imply that Voyager was scheduled for construction over a year and a half prior to the design of the class ship being finished, which seems rather odd.  And finally, as noted earlier, starships do not seem to have NCCs which appear in blocks . . . this would be the logical outcome if you decided to build and give NCCs to several vessel of a class at once.

Thus, with the exception of the Excelsior, we can conclude that the NCC is probably assigned about when a ship's keel is laid.

4.  Shipbuilding Ramifications on Fleet Size in the TNG Era

The problem is, in order to have had 75633 NCCs in the 214 years since 2161, the Federation would've had to have maintained an average of over 353 ships per year . . . an even higher rate in the TNG and pre-TNG eras when you consider that the mid-23rd Century rate was only 20 ships per year.  Either they were going hog-wild with construction and absorption sometime before TNG, or else the numbering scheme was opened up to allow other vessels into it.   The latter seems to be the case.   Consider the Vico, NAR-18834, a non-Starfleet science ship of the Oberth Class, and the SS Raven, NAR-32450, another small science ship only vaguely attached to Starfleet.   Rather than assuming an entirely separate NAR fleet consisting of 30000+ vessels, it is more likely that the numbers are shared between NCCs, NARs, and other N(xx) designations.

That would imply a sudden boost in NCC numbers whenever all those ships were added in, with a more relaxed pace afterward.  However, what we instead appear to see is a series of boosts more indicative of large absorptions.  Later Excelsior registries appear in odd batches, for instance, implying that there might've been several boosts at different times.   There's a batch of Excelsiors in the 14000s, a few in the 38000s, and several in the 42000 range, with some others spread around elsewhere.  Sure, there might've been different batches built years and years later (much like we saw with the construction of the Enterprise in 2245, 35 years after the class first appeared), but given Starfleet's general class-happiness in the 24th Century, it seems odd to think that the 42000 batch would've really been built when the registry would generally indicate.

We can also largely discount the notion of absorbtion of large alien fleets, since of course virtually no Starfleet vessels of alien origin are ever seen.   Huge boost of thousands of vessels would suggest that they ought to show up eventually, but they don't.

Also, given the runabouts, scout ships, and non-Starfleet vessels with NCC-list numbers, one would think that no more than maybe one-third to one-half of the 400 NCC units per year were actually being used by Starfleet for her large starships.   Of course, for all we know they could've been using three-quarters, or even all of those for themselves.  If we assumed that a third of the 400 NCCs were given to large starships, then that would imply a Starfleet starship construction rate of 130 ships per year, leaving plenty of room for Shelby's BoBW2 comment.   Assuming one-half would give us 200 starships per year.

Assuming an average starship lifespan of 50 years, the 130 ship per year figure gives us a standing fleet of 6500 starships, which seems a little low given the evidence from DS9.  The alternative is to either bump up the lifespan (again giving a nod to all those ancient Mirandas, Excelsiors, et cetera) or bump up the construction rate.   If we bumped up the construction rate to 200 ships per year, for instance, we'd arrive at a fleet of 10,000 starships.  If we bumped the lifespan up to 75 years, then we'd have fleets of 9,750 and 15,000 starships, respectively.

In short, the 8-10,000 figure for Starfleet indicated by DS9 seems very workable by the evidence of the registry scheme.  If the average vessel lifespan were 75 years, then to have 8,000 to 10,000 ships would 'only' require a yearly construction rate of 110-130 ships.

III.  Class-specific Comments

0.  The Oberths are frickin' old.  The Grissom was NCC-638, and there are reports that a screen in TNG showed the original Oberth as being NCC-602.   In either case, we're looking at a damned old starship class.  Assuming they were still active by the time of Nemesis (2379), then the class would've been around for something like 190 years, assuming that the NCCs indicate a build date in the late 2180's or early 2190's.  One wonders what a TOS (or pre-TOS) Oberth would've looked like.

1.  The Constellation, NCC-1017, is quite visible in "The Doomsday Machine"[TOS2].   There is the even older Eagle (NCC-956), which comes from Star Trek VI on one of the pages of the "Operation Retrieve" chart.  Even if we discount the Eagle as unreadable background material (which it is), the Constitutions would still seem to have a large registry spread.   Some suggest that we've seen the Constitution herself on various screens with an NCC of 1700 (as per fandom), but, unless there was a great deal of wackiness with registries of that era or an older class of ship easily refit to Constitution specs, the Constellation must indicate an older class ship.  NCC-1700 would possibly be a replacement for the original Constitution, if it was visible at all.

In any event, at least some starships of the Constitution Class may have been around 50 years old by the time of TOS, and, at the time of the Enterprise-A's retirement in 2293, the class as a whole would've been about 80 years old.   This makes the apparent retirement of the class seem a little less odd, though the longevity of the Miranda and Excelsior designs (at least 125 and 95 years, respectively) would imply that the old girls still might've had some life left in them.   There is, after all, a Constitution secondary hull in the wreckage field of "Best of Both Worlds, Pt. II"[TNG4].

2.  I have a star beside the TMP-era Constitution because we do not know what the first ship of this class and type was, or when it was built, besides the 1701-A.   That is to say, we don't know if the refit design originated with the Enterprise herself, or if Scotty's comment that they'd redesigned her simply refers to the extensive refit to the old ship (as opposed to a new-build having a totally original design).   We didn't see any ships with the new engine configuration during TOS (obviously), though we might've expected to during the wargames of "The Ultimate Computer"[TOS2] had they been around.   Other types of vessels (such as the horrendous "Aurora" from the even-more-horrendous "The Way to Eden"[TOS3]) also sported the TOS Constitution-style nacelles during that era.

On the other hand, we do see Spock's warp sled with the basic TMP engine look, and the thing looked like it had been around for at least a little while.  And, we've only seen one or two TOS designs in the TMP era, including a TOS shuttle in the TMP:DE.    There's also a round-nacelled starship in ST3's spacedock scenes.  To the left of center in the cropped screenshot below, you can see the nacelle and engineering hull of one of the study models of Ralph McQuarrie's designs for the Enterprise as she might've appeared in Star Trek:  Phase II.

Given the averages seen above, one could argue that the ship construction slowed at some point between 2245 and 2271 to allow for a lot of refits to take place, which would make some sense.

3.  The USS Lantree was seen (and destroyed) in "Unnatural Selection"[TNG2], but the commonly-given registry of NCC-1837 is only based on the never-seen forward saucer.   Thus, the earliest registry we have is that of the Reliant, as seen in ST2.  Given her registry and my estimates, she would've been built circa 2256.   

The Lantree was the first Miranda seen without the torpedo "roll bar", which may indicate that some Mirandas never had them.  On the other hand, the Lantree had been relegated to transport duty by this time, and would therefore have no need (hopefully) of photon torpedoes and the related launch system.   Hers may have been donated to other Mirandas which had them.

I would surmise that the Mirandas started out looking TOS-ish and were refit or redesigned later.  Alternately, it's even possible that the Miranda Class started life looking just like she does "now", meaning the TMP trend would've started in the 2250's.   However, that seems awfully early.  The TMP trend would've presumably begun sometime in the mid-to-late-2260's (which nicely allows the warp sled to have that aged look, and may indicate some spin-offs from the Transwarp Development Project).    If new-build Constitutions and Mirandas were being built at that time, it's possible that they'd been redesigned to TMP specs.  

I'd also like to comment here on the Soyuz Class, which featured the USS Bozeman (NCC-1941) as a member ("Cause and Effect"[TNG5]).  An obvious derivative of the Miranda Class, we get some firm dating on that ship when Captain Bateson says he's three weeks out of Spacedock and that the year is 2278.   By my estimation above, a new-build of 2278 should've had an NCC in the 2340-2360 range.   Does that present a problem?  Not at all.  As Geordi said, ships of that class hadn't been in service for 80 years, which would imply a retirement date of 2288.   Though we may presume that Captain Bateson had just taken command, there's no reason to assume that the rest of the class was retired after only ten years.  A build date of circa 2260 (as per the registry) seems more likely.  That would give the Soyuz Class ships a service life of at least 28 years, which seems to be more in keeping with the Starfleet standard.  It may even be that the Bozeman had only just been refit to TMP-era specs.  (The class's retirement just a few years later wouldn't necessarily be contradictory, given the similar time between complete refit and retirement for NCC-1701.  Of course, it's also possible that the Mirandas are obvious derivatives of the Soyuz.)

If Mirandas started circa 2255, then the class would be about 125 years old by the time of Nemesis.  While Constitution-fans might balk at the class's longevity versus that of the Connie, the Miranda does have certain advantages.  Her hull is approximately equal in volume, and (not counting the rollbar) her hull surface area is very roughly equal.  However, the ship has a much more compact design, and even in the 2280's the ships were, with rollbar, better armed . . . or with alternate pods available, more functional in multi-role settings.  Dispensing with a single deflector dish in favor of a distributed deflection system may also have been a less expensive option.  

4.  It's been claimed that the Constellation was listed on a screen in "The Abandoned"[DS9-3], with registry NCC-1974.   However, review of the episode demonstrates that this is not the case.  Only the name of the ship is mentioned . . . no indication of class in that episode or her other mention in "Waltz"[DS9-6].   Thus, the earliest known registry is that of the Hathaway, NCC-2593, as clearly seen in "Peak Performance"[TNG2].

Given my chronology above, this would place the ship's construction at or about 2289.  However, the ship's lines, featuring a TOS-angled saucer and peculiar "proto-TMP" nacelles (with their stunted aft end-caps) may suggest that the class itself is much older.   Then again, the ship does have three times the volume of the Constitutions, and only twice the engines.  I would therefore think it unlikely that the ships ever carried a TOS look to them (i.e. round nacelles and such).

Further, whereas vessels like the Excelsior Class Repulse (NCC-2544) were still in service in the TNG era, only the newest of the Constellations with five-digit registries were still in service ("Redemption, Pt. II") . . . vessels like the Hathaway had been retired.  This further suggests that the class may be older than the Hathaway would seem to be, but within reason.

However, with no firm evidence, I am leaving the Constellation Class as a child of the late 2280's.

5.  Extensive comments on the Excelsior are made above, and references to the large spread of NCCs of the Excelsior Class are also referenced.   One unanswered question is when the B-type Excelsior (a la the Enterprise-B) appeared for the first time.  The Enterprise-B was built in the early 2290s, implying that the change was fairly soon after the class ship appeared in 2285.   Further, the registry of the Lakota, 42768, implies that vessels of the B-type configuration were constructed for a long while, right alongside the A-type.   Why this would be the case is unclear, especially considering that both are identified as Excelsior Class starships.  It's possible that the B-type Excelsior was a special-duty vessel.   In any case, its origin and purpose remain a mystery.

6.  Though the Ambassador is identified in the non-canon as NX-10521, this is never canonically verified.  Thus, the earliest registry we have is the Excalibur 26517, as seen on the tachyon grid computer display in "Redemption, Pt. II"[TNG5].   I'm not sure when the Ambassador Class would've appeared, though I'd hazard a guess that it was circa 2325-2335.  The Enterprise-C was lost in 2344, in any case.

7.  The First Contact ships are for the most part a confusing pain in the ass, chronologically speaking.  The Akiras, for instance, sport Galaxy-esque saucers, NX-esque catamarans, Sovereign-esque bridges and escape pods, engines that look like a funky fusion of TMP and TNG, and so on.  

The Saber is a little weird with her saucer-attached engines and hexagonal saucer, but I find her likable nonetheless.   The Steamrunner, however, is just stupid-looking.   So stupid-looking, in fact, that she can only barely justify a single solitary 'cool point' for having a name that comes from the original War of the Worlds.  The only one that looks decent in a chronological sense is the Norway, and that was the one they never used again.

One interesting thing about all of them is that they all share a basic styling of nacelle.  The Norway's nacelle comes closest to the TNG era.

Their registries are in parentheses until I get around to confirming the visibility of them in the film/episodes where they appear.

8.  A large number of Nebula Class ships have registries in the 6xxxx range, implying they're older than the Galaxy Class.  That seems somewhat backwards given the Constitution-Miranda family tree, but it seems to be the case in any event.    If the speed with which the Galaxy Class was constructed and designed (as given above) seems fast, though, the fact that the Nebulas served as forerunner may be of value.

9.  The Galaxy and her registry appear in DS9, confirming the TNG TM on that point.  The ship is also listed as being part of "Star Fleet Battle Group Omega" in Nemesis.

10.  In the chart just above, you'll see the USS Nova, NCC-73515.   She cannot be the progenitor of the Nova Class, since both the Rhode Island and the Equinox had lower registries (72701 and 72381, respectively).   Like the Constitution, one may assume that this was a newer replacement ship.  Given the Danube registries, we're limited to 2368 as the earliest date for the Nova.

The Nova Class seems to have been a technology testbed, in a way . . . she incorporated elements of the Sovereign and Intrepid designs which didn't appear in larger vessels until three or four years later.

11.  As noted, the earliest runabout registry we know of is the Rio Grande, NCC-72452.  The ships are known to have been commissioned in 2368, per O'Brien in "Paradise"[DS9-2].

12.  The Defiant's registry was NX-74205, and Sisko participated in its design and testing as a response to the Borg threat.  This limits it to 2367-2369, and per calculations above I give it 2369.

Given the Valiant's close hull number (74210), we have to assume that the Defiant was built and numbered with a second ship of the class being readied at the same time.

13.  The Intrepid is seen as part of the display on Star Trek: Nemesis.   Voyager's NCC-74656 is only 56 units afterward, implying that Voyager should be one of the earliest Intrepids.  (Indeed, dialogue from "Relativity"[VOY5] indicates that Voyager was to be the first to test the core design on actual deep-space duty, implying that she was indeed the next ship after the prototype, and the first of the class to be commissioned into active service.)  

The Bellerophon, seen in late DS9, had a registry of 74705.

14.  Alas, we've never seen another Sovereign Class starship.   Given the class name, there has to be at least one more besides the Enterprise, but we just don't know anything about it.  In First Contact (early 2373), Geordi said they'd been out in space for a year, so the Sovereign was at least that old.  The launch of the Enterprise-E would thus have occurred during the second season of Voyager, and the fourth of DS9.  The way the registries increase during that time period, the class ship's registry ought to have been somewhere around 75000 (which, incidentally, is what Rick Sternbach says it should be).

15.  The Prometheus could theoretically present a problem . . . the registry that was most visible was NX-59650, on the exterior hull.   That could imply a rather old ship, but one which inexplicably featured the latest tech of the 2370's, such as Sovereign-style nacelles and Intrepid-style hull configuration.   Could we say that perhaps it was just an old testbed that had been retrofit again and again to test new technologies, up to and including Sovereign tech and that multi-vector stuff?   No.  Why?  Because there was a well-identified Nebula Class Prometheus in "Second Sight"[DS9-2], with a very visible registry of NCC-71201.  I find it most unlikely that Starfleet would have two ships of the same name in such a manner.   

The solution is a simple one.  The contradictions are solved by ignoring the hull number and going with the one on the Master Systems Display and the dedication plaque . . . NX-74913.  The hull number may very well have been intentionally false.

Enterprise Averages 

(By "Bond, James Bond" of SCN)

Average size of Enterprises:  469.7m

Average size increase between new Enterprises:  76.42m

(* (G2k's Note) Refers to the unusual vessel first seen as a painting in the TMP Enterprise rec room, as well as on a bar wall in Enterprise.  The length estimate is not canon, and we do not know when the vessel existed.

In TMP, Decker explicitly refers to the vessel as one that was named Enterprise.  With the ring-warp system seen in Vulcan ships, I assume it was either a low-warp gift from the Vulcans, or an early Earth effort that involved a brief abandoning of the nacelle concept.)

Back to Starship Volumetrics