Grid export vs grid spot price

  • 1K Views
  • Last post 13 August 2019
paulatsydney posted this 12 June 2019

Dear Sir or Madam,

I am able to get Gird Export (Actual Load) data from below link:

https://www.emi.ea.govt.nz/Wholesale/Datasets/Metered_data/Grid_export

Let's take Benmore as example. In 2019_04_GridExport.csv, I can find grid export for BEN2202 is all zeros. Which means in this month no generation from this grid point.

Then I go to WITS search for prices for this grid BEN2202 in 2019 Apr, it has prices.

Would you please tell me the logic behind? Thank you in advance.

Best regards,

Paul

Order by: Standard | Newest | Votes
paulatsydney posted this 12 June 2019

Cope grid export data link for reference.

https://www.emi.ea.govt.nz/Wholesale/Datasets/Metered_data/Grid_export/201904_Grid_export.csv

Phil Bishop posted this 14 June 2019

Wholesale prices are produced by running the SPD model. SPD only models the transmission grid, i.e. points in distribution networks that are remote from the grid are not included in the model. There are about 500-600 shadow prices that drop out of SPD but only a subset of these get published by the pricing manager as wholesale spot prices. The precise number changes from time-to-time as assets are commissioned or decommisioned. I think there are currently 243 points at which prices are currently published. See EMI or WITS.

But there isn't necessarily injection and/or offtake at all of the points at which a price is determined or published. For instance, I think there is injection at less than 100 of the 243 points at which a price is published.

As for the Points of Connection (POCs), GXPs, GIPs, nodes, pnodes or whatever you want to call them, and they get called many different things all the time by the Code and the various systems that make the wholesale market work, I assume you know how these things are labelled? The first three characters generally denote a geographical location, i.e. BEN for Benmore, OTA for Otahuhu, etc. The next three characters (or first three digits) denote the kV of the particular part of the transmisison network, i.e. 220 for 220 kV, 110 for 110kV, 066 for 66 kV etc. The last digit denotes the bus or switch (I'm an economist, not an engineer). So I'm guessing BEN2201 and BEN2202 are bits of kit that sit right next to each other in the same switch yard at Benmore. The prices at BEN2201 and BEN2202 would be almost identical, even if not both of them are published.

It therefore wouldn't surprise me at all to find published prices but zero offtake at BEN2202. Try checking BEN2201 - maybe the flow is recorded there?

You may also find the SPD network diagrams helpful.

Cheers

Phil

 

 

 

paulatsydney posted this 14 June 2019

Thanks Phil, you explanation make me know more about the whole picture.

 

While in the file I downloaded for Grid Export data, you can find there is no BEN2201 exists.

https://www.emi.ea.govt.nz/Wholesale/Datasets/Metered_data/Grid_export/201904_Grid_export.csv

 

Best regards, Paul

msouness posted this 17 June 2019

Energy flows are reported as export from the grid, import to the grid.  

When BEN2202 is receiving generation and sending it into the grid, the volume will be reported in the grid import files.

(BEN2201 volumes are reported in the HVDC files)

Regards, Malcolm

icepicknz posted this 08 August 2019

According to the SPD Diagram, SWN0251 connects to SWN2201 to get to the grid.

if I look at the API call for each node

{"interval":"09-AUG-2019 08:15","interval_datetime":"2019-08-09T08:15:00","five_min_period":4,"isDayLightSavingHR":0,"pnode":"SWN2201","load":0.17,"generation":0,"price":172.13}
{"interval":"09-AUG-2019 08:15","interval_datetime":"2019-08-09T08:15:00","five_min_period":4,"isDayLightSavingHR":0,"pnode":"SWN0251","load":3.863,"generation":0,"price":172.24}

It shows the load of SWN0251 being 3.863MW, though SWN2201 reflects only 170kw. Does this mean that SWN2201 results exclude those of connected nodes, or does it mean SWN0251 is possibly connected to another power source?

I have just strated graphing both nodes to see when the Tesla Battery kicks in for generation

Tuong Nguyen posted this 13 August 2019

I am not sure I understand your issue here.

SPD models these two nodes separately and they are connected through SWN_T2.T2. Therefore, the load should be modelled separately.