Tapestry’s Grid Vision, Revisited—Reflections After the Volts Podcast.
Dynamic Line Ratings - that's a low hanging fruit that Tapestry can solve!
After listening to Page Crahan on David Roberts’ podcast, I’ve got a better sense of what Tapestry at X is — and what it isn’t.
Tapestry is not creating a new grid map from scratch. It’s not flying drones or installing sensors. What it is doing is reading existing maps, one-line diagrams, PDFs, and stitching them together to generate insights. Those insights are shared with PJM — or any grid operator who uses their product.
I still don’t know exactly what the Tapestry “product” looks like. But I now have a better idea of what it’s trying to do. As Page described, it’s like Google’s autocomplete — but for the grid. For example, maybe Tapestry can anticipate what kind of report a PJM engineer would build from a solved model. Perhaps it’ll go one step further and pre-run production cost simulations and spit out a draft report.
That’s not a stretch. Engineers today already write scripts — in Python or other languages — to dump results from PSS®E or PROMOD into spreadsheets and generate automated reports. I’ve seen it done. So if Tapestry is going to speed that up and add intelligence on top of it, I can see the appeal.
Can Tapestry avoid these MISO delays?
Source: MISO
Could Tapestry Clarify Who Pays — and Where?
If we’re talking PJM, I’d assume any automated report would include a cost allocation table: DFAX vs. postage stamp. And here’s where things could get really interesting. If Tapestry can help independent transmission developers understand their cost exposure in PJM compared to MISO, that’s useful. I know — I’m mixing apples (ITCs) and oranges (renewable project developers) here. But that’s the kind of separation AI should start to blur. Today, these are still silos.
Right now, interconnection-related upgrades are usually paid for by developers. But RTO-planned transmission — especially those selected through competitive processes — gets paid by load. These are separate buckets. Would Tapestry be able to highlight which upgrades are best handled through interconnection cost recovery vs. regional planning? Could it offer guidance on what’s a “merchant” upgrade and what’s a candidate for competitive bidding?
If Tapestry can navigate that and “throw a bone” to both incumbent TOs and competitive developers — showing who should build what, and when — it might just help us untangle the Gordian knot Page referred to in the podcast.
Speeding Up What We Already Do — But With Better Tools
Page confirmed that the goal is to reduce the time it takes PJM engineers to build models, run simulations, and write reports. That’s the takeaway I had, too. What’s less clear is whether there will be checks and balances in place to catch errors — and whether those checks will be visible to external stakeholders.
Case in point - can Tapestry avoid all this back and forth with stakeholders? Because running the model and getting results does not take too much time, it is the constant back and forth with stakeholders that takes time.
Source: MISO
Model-building errors happen all the time. I remember working with PJM on behalf of a client three years ago. At the time, they were still processing model changes (IDEVs) by email. That’s something I used to do twenty years ago. If Tapestry is serious about automating model-building, I hope it’s helping PJM move past email-based workflows and toward something that scales.
The DLR Opportunity
Here’s where I think Tapestry could really change the conversation in state regulatory dockets: Dynamic Line Ratings (DLRs). One of the most common objections I’ve seen in transmission line CPCN cases is that DLRs are fine for operations but can’t be used in planning.
What if Tapestry could change that? What if it could show that adding DLRs would defer the need for a $100M line until 2028 — or avoid it altogether? That’s the kind of concrete, planning-grade insight that could reshape how regulators and RTOs evaluate alternatives to traditional wires.
Can Tapestry show which of these transmission lines are needed if we had DLRs on existing lines?
Source: MISO
Still Early, Still Worth Watching
So yes, I still have questions. I still don’t know if Tapestry can export a .RAW file or whether stakeholders will be able to shadow-study results. I don’t know whether cost allocation outputs will be verifiable. And I don’t know how PJM’s Transmission Owners will react if the “top 10” Tapestry-recommended upgrades are all 345 kV and above — triggering FERC Order 1000 competition.
But what I do know is this: PJM, Google, and Tapestry are on the right path at the right time. It’ll take a few years to see whether the tools match the vision. But if Tapestry can weave insights across planning, interconnection, and cost recovery — and do it in a way that supports both engineers and regulators — we may finally see some real progress.
Last but not least - don’t forget to look into the textbook I co-authored with Vivek Bhandari and Bill Poppert - (feel free to write a review on the Amazon site if you have purchased already!)
Thanks!