Meaning?

[Jmuccigr]

Thanks for this.

I had a couple of questions about the meaning of a few of the variables, so I thought I would ask here:

1. Is Pumping-consumption not already accounted for in the bulk Hydro number in the balance output? That is, if one wanted net hydro, one would add those two figures in order to subtract electricity used for pumping.
2. I think I understand Self-consumption (the amount of energy smaller providers use themselves before sending the rest to the grid), but is there a way to find out the breakdown of that energy by source (gas, solar, etc)? It's a bit anomalous to have it in there with the source types when it isn't one. I take it from various Terna reports that they count it as part of national consumption.
3. Net Foreign Exchange differs from the numbers provided in the Physical Foreign Flow report via the API. It also doesn't match the Scheduled Foreign Exchange, so I'm not sure what's going on there.

On a side note, ENTSO-E has a more detailed breakdown into more generation types, but their numbers don't match Terna's.

Thanks for any help.

[Jmuccigr]

Some follow-up...

I got the Terna app to play with for another source.

Observations

1. The app's Bilancio Energia matches the energy-balance API. No surprise given the names.
2. The website download page for "Actual Generation" matches both for generation. Again, no surprise.
3. The website download page for "Physical Foreign Flow" (also in this visual on the website) matches the physical-foreign-flow API. No surprise.
4. These two physical flows don't match the energy-balance numbers for "Net Foreign Exchange". Surprise!

I might have thought that they were just presenting different things, but the webpage numbers seem to be ¼ the energy-balance numbers, which might make sense if there were some confusion and somehow the quarter-hour numbers got transformed into hourly ones, that is, they were multiplied by 4 for the balance (and also divided by 1000 to give G instead of M). Here's a quick table for a few times from May 3. Note that I've adjusted for the 1000x in the final difference column:

Time	Website/Flow API	Website x4	App/Balance	Diff
23:45	1484	5936	5.96	32
23:30	1568	6272	6.28	8
20:00	1338	5352	5.36	8
8:00	-81	-324	-.32	4

Notice that the differences are all multiples of 4, which makes me think that the smaller number that appears in the flow API was rounded before being multiplied by 4. In fact dividing by 1,000, truncating to 2 decimals places and then multiplying by 4 would give exactly the number in the balance API, which also has only 2 digits after the decimal place, explaining the smaller numbers that aren't divisible by 4.

It's entirely possible that I'm missing something and being obtuse here, so I wanted to post here before anything else.

The other flow numbers I've found are the "Scheduled Foreign Exchange", here on the website. These are very different, so I don't think they're involved.

[fgenoese]

Disclaimer: terna-py is only a Python wrapper to the official API; it is not a data provider.

That being said, here are some remarks:

• Scheduled Foreign Exchange refers to the scheduled flow, typically related to the commercial arrangements. Physical Foreign Flows reflect the underlying physics, based on Kirchhoff's circuit laws. It is normal that the numbers differ.
• The factor 4 indicates that the two sources refer to different units. Power is measured Watt, Energy is measured in Watt-hours. This means that Energy is the integral over Power. If the interval of the integral is exactly one hour, the difference between Watt and Watt-hours is typically negligible (in these data platforms). When the integral is 15 minutes, you have the factor 4 that you discovered.

I'll add some remarks about the autoconsumption later.

[Jmuccigr]

Thanks. My confusion with flows was what I was explaining in the second post.

Regarding that, I understand the units, but if you’ve got a flow of 800MW for 15 minutes, that means 200MWh of energy transferred, not 3200. Yet the balance (energy) is showing 4x the flow (power), not ¼. Both datasets use 15-minute intervals, too.

Below are screenshots for the physical flow and app energy balance at 23:00 yesterday. You can see that the balance shows exactly 4x the flow with the division by 1000 and rounding up to just 2 decimal places before multiplying (using power units - GW - and not energy, btw):

• 1118 MW / 1000 = 1.118 GW
• round up 1.112 -> 1.12
• 1.12 x 4 = 4.48

BTW If I check France's RTE for their flow numbers with Italy, they show 3380MW for that 15-minute interval, which is about 4x what Italy shows for flow. That is, it more closely matches the larger number, but I'm not sure how reliable that comparison is because the numbers aren't the same even though the trend is.

PS yes, I know you’re not Terna. Just looking for some sanity-check. :-)

[Jmuccigr]

One more:

Italy consumes about 300TWh a year, or 800+GWh a day on average. The units in the generation and balance API reports are "actual_generation_GWh" for each 15-minute interval. If you add those up for balance for May 3, you get 2691GWh, which is obviously incorrect and should be divided by 4 to yield 672GWh, nice for a mild spring day. Generation only for that day yields 2513, or 628 when divided by 4, including self-consumption.

Note that the app says "GW" for energy balance, and that eyeballing the graph above shows an average of about 30-35GW throughout the day, for a total of 7-800GWh, in line with what it should be.

The units in the flow API report are "physical_foreign_flow_MW" and, again, those are mathematically ¼ of the generation numbers for flow in the other reports.

I'm not sure which is correct, though I suspect the larger number since imports are something like 10% of Italy's supply, but something's off.

[Jmuccigr]

FYI I finally wrote this up and sent it to Terna.

[fgenoese]

Feel free to post the reply

[Jmuccigr]

Will be happy to if they reply!