I just downloaded more recent bids (May 2017) from AESO and went through them, to create this live updating graph:

The bids are definitely changing alot, likely due to the increased SGER cost, which you can see on page 5 of this. Output Based Allocation is supposed to come in soon, so I’m not sure how long this will be relevant for.

Generator Bid Sequences

This is going to take up alot of space, but just look through the graphs. I only chose to do graphs for larger units which have multiple bid sequences (more on that to come), whether there is or is not a graph for a particular unit is not important. The title at the top shows which unit the graph is for. The size of the orange/yellow circles is the price that the load was bid in for, on average, during the month of May 2017. The number nearest each circle is the load that was bid in (in MW), in other words, what the generator would ramp up to for the price represented by the circle size. If the circle size is extremely small it means it was bid in for $0. This is commonly done if it is more costly to shutdown a unit than it is to produce free power for a short duration.

How All This was Generated

Currently the bids that each generator puts in are available to the public through the AESO website via this address:

 http://ets.aeso.ca/ets_web/ip/Market/Reports/MeritOrderSnapshotEnergyReportServlet?beginDate={0}&endDate=&contentType=csv 

. The {0} get’s replaced with a date in the format of %m%d%Y (with leading zeroes). This data is available 2 months after the date when the bids were placed. This is to prevent collaboration in order to artificially raise the power price. There was more recent visibility of the bids to the market participants, but that was recently taken away, which you can read about Historical Trading Report Removal. I will compile a month worth of bids, then use R, to take an average of all the bids, using the following code in 1 <- 1
new² <- 1
new³ <- 1
new⁴ <- 1
new⁵ <- 1
plot(g3, edge.arrow.size=.4, vertex.color=“gold”, vertex.size=as.numeric(new), vertex.frame.color=“gray”, vertex.label.color=“black”, vertex.label.cex=1.9, vertex.label.dist=2, edge.curved=0.2) + title(“SD3 Bid Sequences”)

There you have it, you have produced the network graph of the generators bids. The reason why that graph is important, and why I am interested in it, is it has a large effect on the total market merit order. You can see one professionally made at Alberta MSA on page 5. In order to generate the merit order, you must know how much generation is available, and for what price. That means you can only pick one path on that network diagram which was just generated, how do you know to pick the path that ends up in 325MW, or 358MW? Funnily enough you could know because all the bids to higher powers pass through ___MW. This is what makes it challenging to develop the merit order, which is of course done intentionally by AESO, and the reason for the 2 month wait for bid data.

Merit Order Developement

To be continued….