Romulan Singularity Drive and Matter/Antimatter Warp Core Compared: Power, Response Time and Starting Cost Inferred from Intercepted Emissions, 2366–2375
Abstract
No Romulan power core has been in Federation hands for study, and the brief visits of Federation crews to disabled warbirds left no engineering survey, so the performance of the artificial quantum singularity drive can be judged only from outside the hull. We use passive intercepts of warp-field emissions from eleven warbirds (41 sessions, 2,310 observed ship-days) and power-state logs from 14 warbird engagements, both held in Starfleet archives for 2366 to 2375, and calibrate the emissions against 58 passes of Federation vessels with known core output. Cruise output is 1.90 reference units (95% CI 1.56–2.32), about 3.3 times the 0.58 recorded for 19 Federation ships before engagement. Output relaxes toward demand with a time constant of 4.6 hours (95% CI 3.7–5.7), against 62 seconds for matter/antimatter cores. Five cold starts rise with a constant indistinguishable from it (4.1 hours) and take about 30 hours from the detection floor to cruise. Peak output in engagements, 2.47 units for the warbird and 2.38 for the Federation ships, cannot be told apart. The drive therefore reaches its peak slowly. Whether it reaches it at all in battle, or is asked for less, the emissions cannot say. The confined mass, the seed energy and the margin to confinement failure remain unknown. We judge a physical feed limit more likely than a control-system governor, and we cannot exclude the governor.
1. Introduction
Since the 2360s the Starfleet Fleet Telemetry Archive has kept a running record of the emissions that Romulan warbirds leak into subspace as they cross the sensor range of Neutral Zone stations and patrol ships. The record began as a by-product of border surveillance. The Starfleet Tactical Engagement Registry has since added power-state sensor logs from the engagements in which warbirds were met (Starfleet Fleet Telemetry Archive, 2376; Starfleet Tactical Engagement Registry, 2376). Between them the two holdings cover eleven warbirds, a small number by the standard of Federation fleet records and a very large one by the standard of everything else the Federation knows about Romulan ships.
Federation engineers had little reason to study Romulan propulsion until the Romulans returned to the Neutral Zone in 2364. Starfleet has known since 2369 that the power of warbirds comes from an artificial quantum singularity: a black hole of small mass, kept in confinement and supplied with matter. The first systematic analysis of their signatures confirmed it (Halvard, 2372). No Romulan core has been placed in Federation hands for study. Federation crews who boarded disabled warbirds in 2368 and 2369 went to save the ships and not to measure them, and hence every statement in this paper is an inference made from outside the hull. This work extends the Neutral Zone series of the Starfleet Fleet Telemetry Archive and the analysis of Halvard (2372), which first treated the rise from cold as exponential.
We turn those inferences into numbers. We estimate the power a warbird delivers at cruise, the speed with which it can change that power, the cost of bringing the drive up from cold, and the way the drive's own trouble shows itself. We call one ratio the flexibility of a drive: the factor by which a ship can raise its output above cruise in an engagement.
Our vantage is 2376, and no record dated later than 2375 is used.
2. System Description
The Federation baseline is the matter/antimatter core, in which a reaction regulated by dilithium crystals delivers power to the warp coils of the nacelles. Its output follows the rate at which reactants are admitted, and that rate can be changed within seconds (Teague, 2375). We express power in units of P<sub>ref</sub>, the rated continuous output of a heavy cruiser core under the Corps of Engineers convention (Starfleet Corps of Engineers, 2371).
Romulan engineers built a different kind of machine, as far as the outside record shows. Its output is set less by the rate at which fuel is admitted at an instant than by the rate at which the confined mass is allowed to grow and radiate. A simple and rather elegant model follows from conservation of mass. If output P is a fixed fraction η of the feed rate f times c2, and the confined mass M adjusts in proportion to the feed, then output approaches any new demand exponentially with time constant τ = M/f (Osmund, 2371).
During a start the feed cannot exceed what the confined mass can capture, f ≤ M/τ, so the mass and the output grow as et/τ until the feed reaches the level that cruise requires. Once there, the same τ governs relaxation toward any new demand. We take this growth phase as part of the working model, and it is the model's own assumption and not something the emissions show. We adopt the whole as the working model of the paper and not as established fact, because Romulan engineers may have arranged matters differently in ways that no outside observer could detect.
Three sources serve. The first is a set of 41 passive-intercept sessions on 11 distinct warbirds, covering 2,310 ship-days of observation (Starfleet Fleet Telemetry Archive, 2376). Coverage was 58% of the days within each ship's observation window, because ships passed out of range. The second is the set of power-state logs from 14 engagements involving those 11 ships, together with 26 engagements involving 19 Federation ships, whose logs give each ship's output before the engagement (Starfleet Tactical Engagement Registry, 2376). Of the 14 warbird engagements, 11 were confrontations with Federation ships and 3 were joint operations of 2374 and 2375 in which warbirds served beside Federation vessels against the Dominion. The registry does not separate their demand profiles, and none of the three lay outside the range of the other eleven. The third source is a calibration set of 58 passes by Federation vessels whose core output was known from their own logs while the same stations observed them (Brannock, 2368).
The observable is leak flux: the band-integrated subspace emission of the warp field, which depends on the power delivered to the coils. Flux below the equivalent of about 0.0015 P<sub>ref</sub>, or 10−3.1 of the warbird cruise output found below, is lost in the background of every station in the archive (Linwood, 2370).
3. Analysis / Model
Calibration is a regression of log flux on log known output across the 58 Federation passes. The residual standard deviation is 0.105 in log<sub>10</sub> units and the slope is 0.934 (standard error 0.047). Inverting the regression for the warbirds, and averaging ship means so that ships with many sessions do not dominate, gives a cruise output of 1.90 P<sub>ref</sub> (95% CI 1.56–2.32, from a t distribution with 10 degrees of freedom, with the uncertainty in slope and intercept carried through).
The 58 passes themselves had a geometric mean output of 0.62 P<sub>ref</sub>, which is a mean over whatever output each vessel had while passing and not a cruise level.
Inversion rests on an assumption that statistics cannot test, namely that a warbird turns delivered power into leak flux as a Federation vessel does. The leak belongs to the warp field and not to the core, and the assumption is plausible. A difference in field coupling of up to a factor of 1.5 in delivered power, in either direction, moves the estimate to 1.27–2.85 P<sub>ref</sub>, and combined with the statistical interval to 1.04–3.48 P<sub>ref</sub>. We carry the wider band as a systematic uncertainty. Physically, the stations hear how loudly the field speaks, and they learn nothing directly about what feeds it.
Three consequences follow from the model of Section 2, and the records can test the first two. The first is that every change of output follows one time constant, growing from cold and relaxing in battle. The second is that a ship starting an engagement at cruise cannot raise its output quickly. For a ramp of 3.2 hours, the shortest in the record, and a time constant of 4.6 hours, the fraction of the gap to demand that can be closed is 0.50, and a demand of four times cruise would lift output to 2.50 times cruise. The third is that the confined mass is fixed by the product of the time constant and the output, M = τP/(ηc2).
This last relation shows what emissions can and cannot deliver. They fix τ and P, and they fix the mass only up to the conversion fraction η, and if η lies anywhere between 0.05 and 0.4 the mass spans a factor of eight.
4. Validation Against Field Data
Fourteen engagement ramps gave a per-ship geometric mean time constant of 4.6 hours (95% CI 3.7–5.7, standard deviation of ln τ 0.31, 11 ships). The mean fitting error of ln τ within a single ramp was 0.12, well below the 0.31 between ships, and the spread between ships is therefore real. The shortest single ramp lasted 3.2 hours. In 9 of the 14 engagements the rise in output began before the first weapons release, with a median lead of 3.0 hours. Federation ships returned a time constant of 62 seconds (95% CI 52–73, 19 ships).
The warbird constant is thus about 270 times longer (95% CI 207–345).
Five records show a cold start: a rise in emission from below the detection floor to cruise. The mean time from first detection to cruise was 29.6 hours (standard deviation 4.9). The exponential slope of the rise gave a time constant of 4.1 hours (95% CI 3.3–5.1, t with 4 degrees of freedom). The ratio of the engagement constant to the start constant was 1.12 (95% CI 0.86–1.46, Welch-type t with 13 degrees of freedom), and the two therefore cannot be told apart. Under the growth law of Section 2, a floor of 10−3.1 of cruise corresponds to 7.14 e-folds and, at 4.1 hours each, to 29.3 hours. That agreement checks that the rise is exponential. It does not independently confirm the model, since the floor and the slope together fix the duration.
Under the fitted exponential, the output delivered across the visible part of a start is Pτ, or 7.8 P<sub>ref</sub> hours, which is 17% of the 45.6 P<sub>ref</sub> hours that a warbird delivers in a day of cruise. This follows from the fit and is not a separate measurement, and the energy that the start consumes is not visible at all.
The part of the start beneath the floor, where the seed is formed and growth begins, lies outside the archive.
Flexibility separates the designs sharply. The per-ship geometric mean peak-to-cruise ratio was 1.30 for warbirds (95% CI 1.23–1.37, 11 ships) and 4.1 for Federation ships (95% CI 3.71–4.54, 19 ships), a ratio of 3.15 (95% CI 2.82–3.52). The Federation ships' own output before engagement averaged 0.58 P<sub>ref</sub> (95% CI 0.48–0.70), which puts their peak at 2.38 P<sub>ref</sub> (95% CI 1.91–2.96). The warbird peak, 1.30 times 1.90, is 2.47 P<sub>ref</sub> (95% CI 2.01–3.03, wider still under the field-geometry systematic). The two peaks cannot be told apart (ratio 1.04, 95% CI 0.78–1.38). Warbird cruise output is 3.3 times the Federation pre-engagement figure (95% CI 2.5–4.3).
A peak-to-cruise ratio of 1.30 after ramps of at least 3.2 hours implies, under the model, a demand of no more than about 1.6 times cruise. The ratio therefore records what Romulan commanders asked of the drive as much as what the drive could give, and we cannot separate the two from emissions.
The warbird does not hold back more in absolute terms. It must begin to raise its output hours before it needs it. None of the 11 records shows a step in emission that could be attributed to refuelling, over 2,310 observed ship-days within about 3,980 days of observation windows, though gaps in coverage limit that statement.
5. Failure Modes
The archive holds four excursions: episodes in which a warbird's leak flux oscillated about its mean, by 12–24% peak to peak, for two to five hours. The rate is 1.73 per 1,000 ship-days (exact 95% CI 0.47–4.43). In each case the ship cut its output, by a median of 38% within six hours, and then returned toward the earlier level over a day or more. We interpret these events as a fluttering of the confinement that the crew corrected by reducing the feed, though the emissions cannot tell us what fluttered.
Correction speed matters more than its cause. If the time constant governs a reduction as it governs a rise, a warbird cannot shut its drive down quickly. A matter/antimatter core can be starved of reactants in seconds, and when it cannot be starved it is ejected. In the model the singularity offers no such exit, and the logs show no sign of one, and a confined mass keeps radiating while the crew works to correct it.
One record lies outside the archive series. It is the sensor log of the 2369 warbird core failure, reviewed afterwards by the Corps of Engineers (Starfleet Corps of Engineers, 2369), and it shows a distortion in local time near the ship. However, the compilers did not separate the contribution of the failing core from that of a disturbance in the same region, and we do not attempt to. No breach of a matter/antimatter core in the Corps files produced a comparable effect, but that comparison rests on an absence of record.
The method has limits that bear on every figure above. Coverage of 58% leaves room for excursions and refuelling steps to fall in the gaps, and the five cold starts are too few for more than a mean and a slope. Moreover, the 11 ships are those that passed the stations, and the Federation does not know whether they represent the whole Romulan fleet.
6. Conclusion
Emissions put the cruise output of a warbird near 1.90 P<sub>ref</sub>, about 3.3 times the 0.58 of Federation ships before engagement. They show an output that relaxes toward demand with a time constant of 4.6 hours, against 62 seconds for the Federation core, and engagement peaks that cannot be told apart at 2.47 and 2.38 P<sub>ref</sub>. We conclude that the singularity drive is a patient machine, steady at a high level and slow to answer a change of demand. Whether the peak that the warbirds reached was the most they could give is not something the emissions settle. They leave it open.
What remains unknown despite the visits that Federation crews have made to disabled warbirds is the confined mass, the seed energy and the margin to confinement failure. Emissions fix the ratio τ = M/f and the output P, and they leave the mass uncertain by a factor of eight through the conversion fraction. Without a core that engineers can open, the first detectable stage of a start stays beneath the floor, and the four excursions give a rate of trouble but no distance to the limit.
This conclusion has a named rival. The 4.6-hour constant could be a governor, a rate limit written into Romulan control systems to protect the confinement, and not a property of the singularity. A governor set to one value would leave the ships' constants equal up to measurement error. They are not equal, since the constants differ between ships by about a third, which is what a confined mass that differs from hull to hull would give, though the archive does not let us test that against hull size. We judge the physical limit more likely for that reason, and because the constant is the same in cold starts, where no order of battle applies. The rival cannot be excluded, since a per-hull governor setting could not be told from a mass that differs between hulls, and if it is correct the physical response of the drive may be faster than the one we measured.
I thank the station crews who kept the instruments aligned through thin years. The comparison leaves three questions unsettled. The first is the seed, meaning what source of power forms the singularity, and over what time, before the first emission rises above the floor. The second is the margin, meaning how close to the limit of confinement a warbird runs at cruise. The third is the conversion fraction, which would turn a time constant into a mass. Each of them waits on a core that engineers can open. None is likely to be offered.
References
- Starfleet Fleet Telemetry Archive (2376). Passive intercepts of Romulan warbird emissions along the Neutral Zone, 2366–2375. Starfleet Fleet Telemetry Archive, accession FT-2310.
- Starfleet Tactical Engagement Registry (2376). Warbird and Federation engagement records with power-state sensor logs, 2366–2375. Starfleet Tactical Engagement Registry, series RW-3.
- Starfleet Corps of Engineers (2371). Rated continuous output of Federation starship warp cores by class, with the reference-output convention. Starfleet Corps of Engineers Reports, SCE-71-112.
- Starfleet Corps of Engineers (2369). Sensor record of the warbird core failure of 2369 and the associated distortion of local time. Starfleet Corps of Engineers Reports, SCE-69-077.
- Brannock, E. (2368). Leakage spectra of nacelle-field emissions as a measure of delivered power. Journal of Applied Field Dynamics, 8(2), 61–84.
- Linwood, A. (2370). Detection floors and calibration transfer in passive warp-field intercepts. Journal of Applied Field Dynamics, 10(3), 96–118.
- Osmund, K. (2371). A confined singularity as an energy source, the feed-limited response of an accreting mass. Subspace Engineering Quarterly, 17(3), 225–247.
- Halvard, P. (2372). Observed power-up sequences and signature analysis of Romulan warbirds. Starfleet Tactical Analysis Reports, TAR-72-17.
- Teague, M. (2375). Output step response of matter/antimatter warp cores under combat alert. Journal of Applied Field Dynamics, 15(1), 33–58.
Open in Uncited Press →