I suspect such inter-system communication will be driven by market forces and incentives, and as such I don't see a strong market demand for cross-vendor, entire PC solar industry communication. However
- using DOCSIS as a conceptual model, I could envision a grid-tie inverter setup where the grid-interactive portion is controlled by PoCo.
ex., You haven't signed up (or complete process yet) for back-feeding grid, PoCo controls that. etc.
Give PoCo ability to limit backfeed as appropriate (ex. excess power on grid) or request power
Next on the market demand (stability) list might be an extension of above (and as alluded to with request for extra power) would be 'virtual power plant' type capability (ie PoCo requesting at X? price, extra power presumably from battery, though that could extend to shutting off high demand local loads (ex A/C, etc) as same net effect)
- my thinking is that Inverter would receive signal and then 'homeowner' would have pre-configured (at time of system stand-up or entering into VPP agreement/ whenever) response options (let PoCo drain battery, only push battery at certain price to a certain level, etc or whatever the local decision criteria would be. Exact options being local market driven as well as sophistication of Inverter Mfg, presumably with same base capability to 'participate')
The above may be harder, but as they impact entire grid, I'd expect more standardization in that area than others (and I could be really wrong)
Then comes the site specific communication in terms of managing power output from PV, battery, and other (wind turbines, etc). A typical grid-tie situation (backfeeding grid allowed) isn't that complicated with PV generating all it can all the time, and then depending on Inverter settings on battery charge profile, whether to supplement from grid, battery, etc. and the prioritization and timing scheme to follow.
Complications arise when
- not grid-tied (or grid down) and panels producing more than can be used (ie batteries full, if present and no extra loads)
There are existing frequency/voltage adjustments to slow/stop PV production but they do seem a bit brute force type approach to me...
Did I read correctly that Enphase is using powerline (Ethernet over a/c power cables) to communicate with micro-inverters?
though not much incentive, standardizing at least some of that main Inverter/combiner box communication with panels (optimizers/micro-inverters (though would powerline type approach work over DC lines? and this doesn't apply to simple string only setup where nothing to 'talk to'... MPPT/inverter talking to itself)
The next piece would be an Inverter API or similar for work with Home Automation/Solar Assistant type setups (and whatever may come to be in this space). My suspicion is that the market is too new to be focused on standardizing this last step. At this point, I think the onus is on the consumer to demand such communication ability (personally I insist on Ethernet connectivity vs WiFi, but others will prioritize convenience over security and reliability/consistency).
Those are the main 'full system considerations'. Obviously the one I've left out is BMS/battery communication .. though again, maybe a bit early, and with new battery chemistries coming??? is what we have today goo enough (for all but the true die-hard DIYer doing the old mix-n'-mismatch). Is this any different from needing to check memory type and timings with motherboard compatibility?
As for your examples, the issue is that TCP/IP is NOT the only protocol used for network communication, even over the Internet. PCIe for some high-speed, but not all (many others in datacenter/server market). SATA (and SAS) are now legacy interfaces/protocols, etc ;^)
