Watts equal one joule of energy per second. For us, it's a
measurement of how much energy our light fixture is using NOT OF LIGHT
This is why the old rule: "3-5 watts per gallon" can be deceiving, and this rule is only a starting point at best of late. This archaic rule was more accurate when all that was used were T12/T8 lamps which is what this rule is based on.
Keeping this in mind the average T12 has a lumens per watt rating of 40, which means you would need half as many watts of a bulb that produces 80 lumens per watt (assuming PUR & other aspects are equal).
The term "watts per gallon" is getting more archaic with the newer T-5, CFL, the SHO, & especially the new reef compatible LED lights.
Even within LED Lights, one 30 watt LED is not equal to another 30 watt LED.
An example, you cannot compare a 30 Watt AAP/TMC Reef White to a 130 watt EcoTech Radion. However if you were to use an equal wattage of the TMC Ocean Blue or Reef White, you would have more actual useful light energy (PUR) with these per watt of energy used (input energy) than the EcoTech (this is not to say the EcoTech Radion isn't reef capable LED).
More importantly, when you measure input wattage per output of mmol of PAR (which both are easily measurable), you can quickly discern that the Reef White is considerably more efficient. Please read the FULL article to understand why I made this statement.
Expanding a bit more, wattage input of lights versus PAR/PUR output
is where the actual watts used when comparing one light to another is
simply not at all accurate.
Keep in mind that PUR has nothing to do with input wattage. Moreover, PAR efficiency can vary due to PUR, fan use (fans waste input energy/watts), lenses, re-strike (in fluorescent lights in particular), and circuitry (such as daisy chaining of emitters common to many discount LEDs).
For example; the Fluval, Finnex, SB Reef, & Current Satellite are
all Chinese made LEDs that often daisy chain their plethora of emitters
versus LEDs which use PWM, optimized PUR, and advanced
circuitry/drivers with a lower number of HO quality emitters.
The result is a much higher PAR output per wattage input which the numbers speak for themselves.
An example is the SB Reef Light PRO 32. This is rated at 363 watts input energy with a PAR of approximately 881 (100%) at 400mm of air. This comes to .41 watts of input energy per 1 PAR.
Another example would be the Fluval Fresh & Plant 2.0 A3990 which uses 32 watts of input energy with a PAR output about 70 mmol at 400mm. This is .45 watt of input energy per mm of PAR compared to an AquRay NP 2000 at only .08 watt of input energy per point of PAR.