Regarding the idea that there are potentially two cases of a Nebula class and Excelsior class ships having the same name more-or-less contemporaneously – if 23rd century Starfleet has 7000 ships (per Discovery) and Starfleet in the 24th century must logically be a larger organisation, it's entirely possible that different ships have the same name. This might be why the unique registry numbers are so prominently displayed. Beyond them being broadly similar to current, chiefly US, naval ship naming conventions there's never been any hard guidance given on how Starfleet chooses its ship names – whether specific roles, classes, or fleets have specific naming conventions etc.
I broke the known class designs down to four categories: old-age (TMP movie era), mid-age (circa 2300-2350), new-age (2350-start of TNG era), and present-age (TNG-present). I'm not bothering with the TOS era for this chart. I am also putting ship classes in order of when I think they were produced, and any that could possibly overlap into the next category, I listed last: Old-age various Planet of the Titans studies Constitution refit Soyuz various Excelsior studies Constellation Sydney Miranda Oberth Excelsior Mid-age Curry type Raging Queen type Centaur type Ambassador Niagara New-age Olympic Springfield New Orleans Cheyenne Challenger Nebula Saber Akira Freedom Present-age Galaxy Intrepid Defiant Norway Steamrunner Sovereign Nova Prometheus Note that several present-age ships such as the Steamrunner, Norway, and Prometheus classes have registry numbers that would chronologically place them in the new-age category, but are placed in the present-age by virtue of their designs. In these cases, chronological registry numbers don't apply. Anyway, this at least gives us a starting point to be able to assign the conjectural classes into those categories: Old-age Deneva Mid-age Antares Apollo Hokule'a Wambundu Surak Merced Istanbul Mediterranean Renaissance New-age Yorkshire Chimera Korolev Zodiac Rigel Andromeda Bradbury Sequoia Again, that's assuming that registries are chronological. It's just as evident that registries could be based on 'batches,' such as 4XXXX being assigned to a batch of Excelsiors produced in the 2320's, and 2XXXX being assigned to a batch of Ambassadors produced in the 2340's. I'm pretty sure the Hood we see in Encounter at Farpoint was supposed to be the same ship as the 4XXXX Hood we see later in Okudagrams and DS9. The 2541 registry was retconned away.
IIRC, one suggestion that came up was that the Excelsior Melbourne was due to be retired in favor of the Nebula Melbourne, and this is the one that would have been Riker's if he accepted command. The Borg assault nixed that plan and both ships were lost in the battle. I personally have long favored the batch concept for registries, because I consider it far easier to manage than trying to make them all a single chronological sequence. It also makes it easier to work with sources like Jackill in which different classes might have similar batch numbers, and refits of a lead ship could result in a new class registry (as happened with Sisko's Saratoga, which for Jackill was the lead ship in a new variant. Presumably it was the same Saratoga from STIV that was modified).
@Dukhat, your list is pretty good, and I won't get into all of it, beyond not entirely agreeing with non-sequential registries. Only thing I'll say on that is that registries are chronological to when a ship is authorized -- but not necessarily built. For instance, the Galaxy Class Development Project was authorized in 2343. Various subassemblies and systems were designed, reviewed, tested, and approved over subsequent years, but the Galaxy herself didn't have her "keel laid" until 2350. And she wasn't launched until 2356. I'm figuring the first dozen authorized Galaxy-class ships got their registries assigned when the project was started in 2343, but -- beyond the lead ship -- probably weren't named until later in the process. Like, there was probably discussion in the halls of power about whether to name one of these new super-vessels Enterprise -- and then the Enterprise-C was lost and it became a no-brainer. But those thirteen years between authorization and launch of this new class are full of additional authorization and building of proven designs that are faster to build. Which isn't getting into how I flat-out headcanon-erase the Steamrunner and Saber classes entirely, due to ILM designers not understanding the tech they're working with. "I recognize that Mike Okuda made a decision, but given that it's a stupid-ass decision, I've elected to ignore it." I go with him on a lot of things, but also don't hesitate to call him out on failure to properly research. Given I only just this past decade ran across the pics of the Excelsior miniature being built, and being relabeled as the Hood and Repulse, prior to its TUC re-work, given Mike wasn't over at ILM on the visual effects stage and it wasn't even his department, and given that the registry wasn't visible on the screen in the case of either ship and we only got the Repulse's registry from her shuttle... I wouldn't be at all surprised if Mike never even knew the miniature had been relabeled, and felt free to make up a registry number for this (he thought) unspecified ship. A curious inverse of the Yamato incident. I'd be cool with that if they didn't have the same, visible, registry number. This... is a longer thing that I want to get into in this thread. I've danced off onto the soft shoulder of the topic a few times. I'll just sum up that Matt Jeffries had notions in mind that made it to the screen in TOS that numbers were assigned in blocks, and, later, that the prefix indicates type as well as national registry, so that one could understand what they were looking at from breaking it down at a glance. Our main hero starship for the obligatory example: 'N' is the broad indicator of Federation registry*, CC means it's a Space Cruiser/Starship Type, '17' indicates it's Starfleet's seventeenth Cruiser class, and the '01' indicates the first production hull built after the prototype. [*Okay, I'll take up some space to put in this much expansion: The doubled letters indicate an active Starfleet vessel. CC for Cruisers/Starships, and the only other one Matt specifically conjectured was DD for Destroyers. I'm happy to hold it there -- there's too much on-screen indication that Starfleet's main distinction is between independent multitaskers [Starships] and more single-focus or fleet-dependent "smaller vessels" like Destroyers/Escorts, science vessels, and transports. Here is a link to my big long post on it, and here is my follow-up clarifying my take on Frigates in Trek a little.] For... reasons... By the time we got to the TNG era (in the real world, i.e., circa 1986), registries were just applied sequentially regardless of class or mission profile. Rather than try to take one model and forcibly and imperfectly superimpose one model on the other, as I've seen others done over the years, I prefer to say that the one gave way to the other, for in-universe reasons that can be come up with and that will make perfect internal sense. So, by the 2300s, blocks still occur, but as an accident of bulk-authorizations, rather than an indicator of when they're actually built or what kind of ship they are. My personal take of the 14XXX block of Excelsiors is they were authorized in the panic after the Tomed Incident, and the 42XXX block in response to some crisis or conflict in the 2330s. **EDITed to add links to my posts elsewhere on here in my footnote on Jeffries' registry stuff.**
Very early in the first season of TNG, it appears that the newest ship registries, as of whatever year they thought the show took place in, was 5XXXX, before it was later bumped up to 7XXXX. This is evident in an early shiplist with the U.S.S. Galaxy having a 5XXXX registry, and the NCC-53911 Tsiolkovsky's dedication plaque with a commissioning stardate of only a year before the Enterprise-D's (ignoring the fact that the VFX guys ended up using the Grissom Copernicus NCC-640 filming model, as I'm guessing Okuda, who made the plaque, was assuming that a new model was going to be built for the ship.) Since we also see other older ship designs with Galaxy class attributes (the New Orleans, Cheyenne, Springfield, Challenger, and Olympic classes) with 5XXXX registries, you're probably on to something. I strongly dislike all four of those FC ships, but mainly because of the low registries those ships were given when they were obviously meant to be new and contemporary with the Enterprise-E. If their registries were all 7XXXX I wouldn't have had as much of a problem with them, although I'm still not crazy about their designs. But I truly believe that ILM never meant for those ships to be scrutinized up close, and that we wouldn't see them again after FC. That's also why Lower Decks annoys me, since apparently the people in that show who are in charge of giving ships registries have their heads up their asses. What I'd like to know is why he thought NCC-3XXXX-4XXXX was in any way a good idea for an Excelsior's registry number. The 1XXXX numbers I get. But 4XXXX? And don't even get me started on the whole Ambassador class screwup...
Hey, that reminds me of the Prometheus in Voyager... I like the Akira and Norway -- agreed, registry issues aside -- but in the case of the other two... *sigh* Matt put the Enterprise's engine way out away from the habitable volume of the hull for a reason. He figured that, regardless of not knowing specifics, whatever field they generated to warp the fabric of spacetime, however much power was needed to do it, and whatever was needed to generate it... it probably was none too healthy for people to be near. The sets make it problematic to determine, but I feel the Jeffries tube Scotty's always crawling up into to fix things has to be angled for a reason. My hot take is that there's one at each nacelle pylon root, and probably another below the anchor point for the interconnecting neck. I find it telling that, to fix something engine-related, he never goes farther up. FJ put bulkheads in the pylon/engine access for each stage of safety requirement. Something similar would probably be in play with the TMP ship. In TNG, we can see the advance of technology. Andy designed docking ports into the nacelle pylons right below where the engines attach, presumably for systems inspections. But we've also seen that the nacelle itself is crewed during normal operation, force field and materials sciences progressed to the point it's safe to be up there without a rad-suit, even while the engine is operating. Have to power the engine down entirely, though, to be able to be in where the coils and injectors are. So to see the Saber and Steamrunner (which isn't even a word, dammit! They called up the old "streamrunner" term but didn't like how it sounded so they dropped one 'r' -- that's not how it works!) with the warp engines anchored directly to OR RUNNING THROUGH the habitable volume of the ship makes me stabby. The oddball solo registries? If Starfleet wants a good, capable Cruiser without the resource demand of an Ambassador, it makes sense. Roll one out to fill a billet need. Of course, I definitely feel the look of the class should have evolved over that time-frame as new tech came along. I'd love to see a 24th-century Excelsior with phaser strips, for instance. Maybe short, Ambassador-style oval saucer windows down near the rim of the curved saucer top. Maybe an updated dish like the Excalibur's, stepping stone toward the Galaxy dish... In the "finished Excelsior sketches I've done, I give the older ships straight angled pylons like the Enterprise class, albeit at a shallower angle. Then, for later ships, I give it the "L" shape the prototype had. Something something, same thing with NX-10521 (Ambassador), mumble figured out some benefit to that pylon shape that had eluded engineers before with the Excelsior, blah blah new standard until Intrepid. Which? I can think of several.
Steamrunner's design layout resembles Springfield/"3-nacelled Excelsior Variant"/Oberth. I can buy it as a scout. When I watch the show and try not to think of deck plans and such, this is what makes sense to me. Now for a lot of info I was needed to look in my notes for: Sternbach's six big classes are explorer, cruiser, cargo carrier, tanker, surveyor and scout. For comparing these classes wartime/present day/video game ship types, but using only existing Starfleet ships from the show, I got this(with some examples): Scout=Scout--example: Oberth, Springfield, Steamrunner Surveyor=Escort/Destroyer: Defiant, Norway Tanker=Frigate: Constellation, Cheyenne Cargo Carrier=Frigate or Shuttle Carrier: Miranda, Akira Cruiser=Cruiser: Enterprise, Excelsior Explorer=Battleship--example: Sovereign I also used Memory Alpha's then-exhaustive NCC list to match unseen-named classes with unnamed-seen classes (and classes for which we have proof there was a model but no proof they were really onscreen. I made some assumptions about NCC numbers that not all fans will want to use, but this work took many hours and I do not want to redo it at this point. The resulting list makes sense and allows a person to interpolate the classes of some ships that are unknown. 1st assumption: the first two digits are a build order number, which could but might not always indicate class. So NCC-47321 would be part of build 47. 2nd assumption: The last 2 OR 3 digits indicate a serial number within that build, so NCC-47321 is the 321st ship in build 47. --Important to note here is that, in this system, the next ship after 1299 might be 1300 if a new build is started, or 12100 if the same build continues (some fans are going to hate this part, lol). This solves the problem of enemies knowing how many ships Starfleet has, and, to me, helps alleviate the issue issue of the 30000's Mirandas and 40000's Excelsiors. There are surprisingly few 3000 and 4000 ships in general, so this helps with the distribution of numbers. 3. When interpolating a given class name was very fuzzy even with the system above, I used a logic suggested by a poster here: NCC-1664 Excalibur is replaced by NCC-1764 Defiant is replaced by NCC-1864 Reliant. So using this and the NCC chart as a guide, these are the matches I got: Apollo--either USS Jupp or ringships from Unification Bradbury-Voyager Prototype Hokule'a-Excelsior prototype (the one that looks most like the final) Korolev-4 nacelled "proto-Nebula" Mediterranean-Ethier unseen Pegasus design, or real name of class that includes Centaur/Curry Merced-DOES have unique model combining Miranda and Galaxy Parts Rigel--Enterprise-C variant of Ambassador Renniasance--Enterprise-B-type Sequoia--USS Elkins Wambundu-4 nacelle Excelsior prototype Zodiac-USS Yeager from DS9 (assumes registry of 65674 was really 85674) I can add more once I set out my notes better. EDIT: If I really want to
I admit, while I understand the criticism of designs like the Steamrunner in relation to the nacelle design (compared to how Matt Jeffries envisioned the nacelles being away for safety reasons), I also think it and related classes like the Defiant are nice designs aesthetically. It could that by the TNG era, technology had progressed to some degree so it was possible to build some classes without necessarily needing to jettison nacelles in the event of an emergency. Just shut them down, while giving the crew more ready access to their systems for lesser repairs. We've also seen a number of alien ships like the Galor that don't seem to share this philosophy, and have their engines built directly into the hull too. So I would say it depends in part on the mission profile and the resources available, although a design with separate nacelles would still be more common.
I've warmed to the Galor over time....looks rugged. Excelsior may have the best warp dynamics as part of its use of transwarp. One could say the NX Excelsior was scaled down--and Sulu's ship larger, if you keep to the bridge.
Star Trek registries are only vaugely chronological. 5xx, 6xx, 10xx, 13xx, 16xx, 17xx, 18xx, 19xx, 20xx, and 21xx are contempoarary (and 38xx if you include the tug class). Also, of the TMP era ships, the Constellation is the most advanced in many ways and is not older than the Miranda or Soyuz classes. It also has a higher registry sequence indicating it was later. There seems to be a change from before 2000 to after 2000. This happens later with 1xxxx marking the Ambassador era, 3xxxx and 4xxxx showing up all all sorts of old hulls, 7xxxx on new looking ships. But beyond that the numbering is random. I reconcile it by the Federation authorizing construction with NCC and then Starfleet builds what it need of what was authorized. For some classes they build all, for others they don't. So in the end it is an weird mix that the writers and graphic artists made up on the fly.
I have relented about 2 of the study models. They were on screen with enough clarity and they fit in with the flow of ship design. The others might as well, but these in particular. The first is the 1st 4 engine Excelsior study model, also labeled Excelsior NCC 2000. These warp nacelles greatly resemble the Grissom. So the belong to the same design time frame. If you take the behind the scenes reasons for the existence of the model and layer that into the Trek universe, you get a 3rd Transwarp prototype. Pulling the name and registry from the Ingram plans, I'm labeling this the USS Simanov, NCC-2002. It appears to be made from an AMT Enterprise saucer. So I theorize that while the third one started, it was the first one completed and it was a failure in terms of faster speed. The design theory led to the Constellation class (USS Constellation was active with its NX registry in 2293). This image also has one of the older Enterprise study models. This would be the TOS equivalent of the USS Discovery. Again, experimental engines. I see a pattern developing in Enterprise and the TMP movies. Lots of experimentation with making warp drive faster. It seems to be a Starfleet obsession. This would lead to building a lot of ships in low production runs looking for the next successful jump in warp power. These would lead to a larger variety of ships, none built in large numbers and probably a lot of them one-off designs. Starfleet found several designs that work and put them into production in large numbers to be the workhorses of the fleet while they poured more money into pushing warp technology. But after the Transwarp project things seem to stagnate for a long time. The warp factor scale changes. I see this as part of why there are so many Excelsior, Miranda, and Oberth class ships. They had some winners for quality and they built them. The US military did this with several aircraft. The F-4 Phantom was everywhere, in every branch of service and used world wide. At other times there has been more variety. In WWII there was a lot more variety, with various planes specializing in different roles for different branches of service. So Starfleet went through this in the TMP era, building a lot of Miranda, Excelsior, and Oberth class ships that ended up serving, continuously or sequentially, for over a century.
The other area to look at is versitility. We know the Excelsior and Oberth classes did not see many varients, but the Miranda certainly did. This gives us quite a variety of uses for one class of ship. That sort of adaptability is what ends up giving this class such a large number of on screen appearances. So the Miranda design is not as new, but it is more versitle and can fit more roles than other designs so it surrvives a long time because of that.
Are the engines supposed to move on that model? Because if so that is badass Makes you wonder what TNG/DS9/VOY-era ships would have ended up looking like if the Excelsior had had this arrangement instead of a more traditional saucer-secondary-twin nacelles layout.
There's a lot of fan variants of the Oberth which are at least halfway plausible, even though we've never seen them on-screen. They are based on the idea that the Oberth is really just the upper part of the ship and the secondary hull is a swappable module. The default Grissom-style variant we usually see is carrying a big unmanned sensor array and automated probe launch system. Other variants include a Ptolemy-style tug, a tanker, and a detachable hospital station module. My particular favourite is the Jester, which swaps out the big sensor array the Oberth usually has for a Miranda-style photon torpedo launcher roll bar. *Edited for formatting errors.
Good point on design versatility. A secondary hull with attached engines on pylons and a centerline neck attachment to the saucer gets in the way with multi-mission specific modifications as shown. This is why the Constitution Class ships were retired and the Miranda Class ships were used for another hundred years.
Regarding the comparison between Excelsior and B-52.... Wikipedia indicated that the B-52 first flew in 1952, and entered service with the USAF in 1955. The plane was upgraded 2013-2015, which by that point the design had been in service 60 years with the same operator. Reasons given for longevity are high subsonic performance with relatively low operating cost. It is expected that the B-52 will remain in service into the 2050s. I like to compare the Mirandas with a smaller and somewhat older plane-the DC-3. According to Wikipedia, the DC-3 first flew in 1935 and entered service in 1936. And the DC-3 is still in service, after eighty-five years. I wouldn't be entirely surprised if there remain a few DC-3s in service in the year 2036.
Except neither of those latter two have the warp engines running through the ship. I'd flip your Scout and Surveyor examples. Scouts are basically Destroyers with better sensors. Surveyors are more dedicated non-combatants like the Grissom in TSFS. What's more, his list was never laid out onscreen. Thanks to DS9, we know 'Escort' is a category, but not one that Rick included in his list. I'd also argue that cargo carrier and tanker are more mission profiles than dedicated starship types. We've seen a Cruiser -- the Lantree -- running as a cargo ship. We did see dedicated Starfleet freighters back in TOS/TAS, but by the 24th century, they just seem to be repurposing older starships for that role, instead. We also have Frigates called out in the first season of TNG, and also not included on Rick's list. A revised list should be something more like: Explorer, Cruiser, Escort. I've done some deep delving into the classifications thing over the last several years and -- taking canon and fanon into consideration -- treat Frigate as an accepted shorthand for either a Cruiser that has been unbalanced toward combat or an Escort/Destroyer that's been given additional equipment to balance it out somewhat. If a base Miranda has no roll bar (like the Lantree) and is a Light Cruiser, the upgunned Reliant is a Frigate. If the single-nacelled Saladin-class Destroyer is fitted with a second warp engine nacelle and some science labs and better sensors, it becomes a Frigate (basically a Loknar or Wilkerson). Frigate, Scout, Destroyer, Surveyor, Cargo Carrier, and Tanker are all more apropos as mission loadouts, subclasses, variants, or similar alterations. Snipped for brevity (a losing battle with me, I know). Your methodology is sound, and if it works for you, that's the best anyone can hope for. I personally disagree with your conclusions, because I manipulated the same sparse and too-often contradictory data toward different conclusions. I'll spare you the process and try to thumbnail my conclusions: • Jeffries' system, if fully realized and used by TPTB in TOS, TAS, and the first several films, would have been in use from the beginning of the unified Human Starfleet in the 2130s up until the mid-23rd century. Human -- and, later, Federation -- shipbuilding capacity was such that most classes didn't come close to hitting the hundred-hull cap each number block tacitly imposed. As indicated by the "Court Martial" wall chart, obviously this was starting to no longer be the case by that time period (NCC-1697). Also to go with Jeffries' unrealized notions, Starships/Cruisers have NCC prefixes, Escorts have NDD. • Around the time of the TWOK/TSFS/TVH triptych, expanding shipbuilding capacity and the non-Human races participating more in Starfleet necessitated a re-work of the vessel registry system. Starting with NCC-2500, all registries were assigned, regardless of class or type, in sequential order ongoingly. Any apparent "blocks" would now be due to batch orders being authorized. The NCC prefix now just indicates an active Starfleet vessel, with NDD dropped. No more distinction, per Cartwright in TVH, between "starships" (Saratoga, Yorktown) and "smaller vessels (Shepard). NX is still for Federation experimental starships, nearly always (allowing for the odd exception to show up) created by Starfleet, they having the most resources for research, development, and construction. • This was the approach that was able to preserve the most material seen onscreen intact. The only things that need changing are the prefixes of the Revere, Columbia, Grissom, and Jenolan to NDD, and the shifting of the registries of the Bozeman to 1841 and the Constellation (the class lead ship, not the TOS one) to 1900. Any other approach I tried breaks either the Jeffries or Okuda system to some degree. Not even remotely a problem. All the Yamato's registry of 71832 tells anyone is that Starfleet has ordered at least 71,832 ships -- nothing in that number says how many are currently active, or how many have been built since. Overall not bad. I'm not thrilled with a Human name being given to a Vulcan ship class. I personally have the Apollo class as the Phase 2 study model seen in the Qualor II yard, and the T'Pau was not that class (Surak works nicely, though, as we haven't seen it, either). The Centaur I gave the unseen Chimera class monicker. I disagree with the Rigel andRenaissance interpretations you have. Definitely like your revision of the Yeager. There's a lot of unknowns there -- how do those other races' FTL drive systems work? How much higher are the levels of harmful substances and/or energies in the habitable volume, because of racial prejudices (Klingons be all, "What, you can't take a little radiation, you weak pa'takh?")? How much efficiency loss do they accept to reduce engine vulnerability? Etc. Leaving out that I treat the Constellation's registry as an error, the Republic was a training vessel when Kirk was at the Academy. I wouldn't really call it a contemporary of the Enterprise as far as active service during TOS goes. 15 and up, though, yes. Bearing in mind that, at the time of TOS, the Enterprise had been in service for over twenty years, and the hull number had been assigned about twenty years before that. Forty-odd years is a decent amount of time for the Miranda, Constellation, Excelsior, and Federation classes to come along after, with their own number blocks (and the Belknap at 2200, Enterprise at 2300, and Menagha at 2400, prior to the big registry revamp). I actually just handwave it as 1800 and make the cargo-pod-hauler a Miranda subtype. The unreadable text we see on the TMP/TWOK screens I ascribe to the Surya class Escort. Yes, I play fast and loose with some of the fanon designs, especially if they can be made to fit the canon with comparatively minimal tweaking. While the Excelsior's hull number (and probably a handful of additional vessels at the same time) was authorized circa 2265, the ship itself wouldn't be launched for a couple decades. While I figure some classes benefited from the tech coming out of the Great Experiment development project, at least earlier on the new classes coming out would be extrapolations, refinements, and upgrades of the Constitution/Enterprise design generation. We'd probably start seeing Excelsior-derived designs leaving dock starting around the time of TVH (point of fact, from the angles of the models involved, there were two Excelsior-class starships in Spacedock when the whalesong probe knocked it out). It doesn't really seem to have an angle to the saucer edge. Plus no "blister" on the saucer top or undercut on the saucer bottom. Rather than cut off and fill in those features, they probably just took some foamcore, cut a circle, added the superstructure and extensions, and bondoed and painted it. The only recognizable kit parts to me are the Phoenix missiles used in the engines. Why? My takeaway from TOS is that, since several alien entities propelled the Enterprise at heretofore-never-experienced Amazing Blinding Speeds™ (warp 14 -- old scale -- at one point), Starfleet, having seen it's possible, started trying to figure out how to do it without outside "help". Rather than an obsession with faster for faster's sake -- they'd seen it could be done, so the rest was just an engineering challenge. The Constellationclass seems to have been an early attempt. Premise probably being "let's just add more engines". The result was what Picard describes as "under-powered". Probably no faster, but with better endurance (the Stargazer and Hathaway were waaaaaaayyy out there). I can see that quad-Excelsior being an attempt to do better starting with better engines, maybe launched sometime in the late 2270s or early 2280s -- between TMP and TWOK, but more probably after the Khitomer Conference. As for scarcity, lack of evidence is not evidence of lack. In "Unification", no one on the bridge pointed to the Phase 2 study model and said "Hey, there's that thing -- they only ever built the one". So we have zero data from which to derive production numbers. If you want to say it was a one-off with experimental engines, cool. My personal take is it's the Apollo class, predecessor to the Ambassador class, and extant in decent but not huge numbers. Neither of us can claim to be right, though, based on what we know from onscreen evidence. The Lantree and Sisko's Saratoga I feel are base Mirandas (albeit one with mission-specific outrigger sensor pods mounted). We saw enough Reliant types that that is definitely one subclass. The "Soyuz class" is another. The Antares is a 24th-century update of the Reliant subtype (same phaser cannons, new torpedo pod). But the Bradford is different enough I consider it its own class. If it had the same number of engines in the same place as the rest, sure, but as it is, no. Be like saying the Enterprise and Saladin are variants of the same class because they have the same saucer. The Excelsior and Grissom have at least as many variations. Sure seems like it. I like to think it worked, but not as well as hoped, and it wasn't dusted off again until the Intrepid class used it to better effect a century later.
There's no reason to assume that there's conventional nacelle interiors between the Bussard collectors and the nacelle proper going through the body of the Steamrunner. It's plausible that their warp coils are only in the aftmost two thirds of the twin booms projecting back from the saucer, lining up with the warp field grilles. Jeffries' original explanation for making the nacelles discrete, separate components (with self-contained power sources and fuel reserves) has emphatically not been supported by everything we've seen about warp drive technology from TMP onwards.
I'm not even talking about that. Reactors and power delivery is irrelevant. Sure, the warp coils can be all in the bit sticking off the back. I'm talking about this acknowledged high-energy-EM-field-generating bit that pulls in interstellar matter right next to windows.