kW : kilowatts POWER - think of a 100 Watt bulb which needs that much instantaneous power to stay lit. This is the "main" number how residential Solar systems are rated (also called nameplate capacity) and is the (semi) optimal DC output (I won't get into DC/AC ratios, etc.). So for instance, eighteen 400 Watt panels would be a 7.2kW system ... so this is also used in pricing. BTW, one thing that surprised me was that the pricing from all vendors was almost linear - i.e. it didn't matter if you went with a small or large system, the dollars/kW was the same from each respective vendor. In fact, for Tesla it was exactly the same $2.57/kW on all quotes from 5.2kW to 10.8kW. This actually makes the economic analysis easier, since we don't have to factor in scaling considerations.
kWh : kilowatt-hours ENERGY - that light bulb needs a 100 Watts of power and when it has been on for 10 hours, that's a 1000 Watt-hours (or one kWh) of energy. Since this is what you consume, it is what really matters. So over a time period, a solar system will produce some kWh of energy, which is a function of the size of your solar array (in kW), how much sun they get (varies depending on time-of-year, location, weather, orientation, shading, etc.), efficiency of your panels (less in hot weather and they degrade over time, and if dirty), efficiency of your wiring/inverters, etc. An easy example is the same 400kW solar panel in sunny Colorado will produce more kWh of energy than one in not-near-as-sunny Seattle. Historically, my total electrical energy usage for a year is about 7,500kWh .. so that's how much ENERGY I need ... but I also need to match my POWER needs - i.e. I obviously pull less kW at 1:00AM versus cranking the A/C in the heat of the afternoon ... so there are time-of-use variations. This is why solar (alone) will NOT power your house (LOL especially at night!) so you would need some form of energy storage - i.e. a battery. And you are going to need a BIG battery(s) if you hope to run any major appliances for an extended time.
Capacity Factor - This is the ratio of the actual energy output (kWh) of a solar plant over a period of time compared to its maximum possible output if it had operated at full nameplate capacity (kW) for the same time period. So for instance, I have a 7.2kW system ... so if it produced that full 7.2kW for 24 hours a day for an entire year, that would be 63 kWh of energy. However, the expected production output of my system for a year is just over 9kWh ... so divide those numbers to get a capacity factor of 14.5%. This should NOT be surprising ... since sometimes the sun doesn't shine (DUH!) and I actually rarely see more than 6kW output since that 7.2kW is based on perfect conditions.
Solar Cell Efficiency - As of 2023, most solar panels are about 20% efficient ... which means that they convert 20% of the sun's light into electricity. Sure, it would be nice if it was a 100% ... but so would Fusion power! There are more efficient solar cells, but you have to balance out cost, longevity, etc. For instance, if you are putting solar on a space satellite, you would be willing to pay a LOT more (for efficiency) than a home owner! ;-) NREL has good historical chart showing how they are (slowly) getting better. A recent technology (Perovskite) provides a good jump in efficiency (and is still cheap to make) but their expected lifetime are significantly shorter than the 20-25 years of silicon solar cells. As with anything, there is a BIG difference (and leap to be made) from something that works well in a laboratory to scale up for large-scale and long-time deployment.
TOU : Time of Use pricing - you pay more depending on when you need it. For instance, peak electrical demand (in kW) is often on a hot, sunny afternoon when everyone cranks the A/C ... so it's not surprising that electricity is more valuable then ... *especially* if it's get cloudy (reducing home/utility solar output) and the wind is not blowing. I was actually one of the first people to get this as I signed up when Xcel offered it as a Beta program. It's now being rolled out everywhere. Clearly it would be advantageous if people "time-shifted" their electrical usage to reduce the peaks, but that is easier said than done (unless you have a "power sink" such an EV or house battery) and this can lead to overall higher energy consumption because you might do something like "pre-cool" your house in the morning.
Battery Backup - So it should be obvious that batteries are primarily rated by how much energy they can hold as that is the most important parameter. For instance, the Tesla PowerWall (as of 2024) is 13.5kWh ... so that means you can run a 100Watt light bulb for 135 hours ... but my 4-Ton, 14 SEER Air Conditioner would only run less than 3 hours! A secondary rating is the discharge rate - how much power it can put out ... and for the Tesla Powerwall, it's 5kW continuous and 7kW peak. So if you try to run your A/C from the battery AND then plug in your EV car, that's going to pull more power than the battery can put out. Batteries are quite expensive - adding ONE would be an additional $8,112 after 30% tax credit! So sure, when utility power is out, you can have the lights on ... but don't try to do anything significant for an extended time. BTW, almost all non-battery solar systems will NOT provide energy when utility is out ... unless you have a transfer mechanism ... since you don't want to back-feed the power lines. So the logical use case for batteries if you have a medical device (or something) that absolutely, positively must remain on 7x24. Another corner case is arbitrage where this is substantial difference in TOU pricing ... so you charge your battery when utility electricity is cheap (typically overnight) and "sell" it back when demand is high and they really need it - typically when lots of demand due to A/C and drop in overall solar production in the evening.
Bi-Directional Charging - This refers to using your EV car to not only charge from your house (via utility or solar) but also be able to send that power TO the house/grid. It actually makes a LOT of sense ... since some EV Cars have 10X the battery capacity of a Tesla PowerWall - the extended-range Ford Lightning is 131kWh. If you already have an EV, definitely something to look into as there are some (expensive) solutions that allow this. I think it will take about a decade for this to become cheap/common.
NEM : Net Energy Metering - this is how your utility "settles" up with you for the excess energy you generate. You hear a lot about how electrical rates influence the economics of Solar, but I'm increasingly convinced that NEM will have a bigger effect - more about NEM here.
Utility Power Projects - When you read Press Releases about these, pay close attention to kW, kWh, and capacity factor. For instance, traditional power plants (especially nuclear) have a capacity factor close to 100% ... whereas wind/solar are much, much less. So for the later, PR folks often specify kW (or MW/TW) but that is capacity, not production (kWh) which is much less versus a nuclear plant. Plus there are timing issues - i.e. your solar won't produce much power at night when you want to turn on the lights! ;-) Obviously batteries can help here ... but they need to be BIG. I.e. it's one thing to provide instantaeous power (if something trips offline, frequency stabilization, etc.) but quite another to do for multiple hours/days. So similar issue as discussed about the Tesla Powerwalls ... as a 2,000kW battery system may be enough to power a thousand homes, but depending on how much energy is stored, that may only last for a couple of hours - much less if people use their A/C, etc.