Free Destination Charging: 50x L2 80A Stations @ Caltech, Pasadena, CA

[chuck-h]

Separate conduits for communications vs. power wiring. This is for inducted current reasons as well as overall voltage safety.

800.133(A)(1)(d):
"d) Electric Light, Power, Class 1, Non-Power-Limited
Fire Alarm, and Medium-Power Netvvork-Powered Broadband
Communications Circuits in Raceways, Compartments,
and Boxes. Communications conductors shall not be placed in
any raceway, compartment, outlet box, junction box, or similar
fitting with conductors of electric light, power, Class 1, nonpower-
limited fire alarm, or medium-power network-powered
broadband communications circuits."

Does "communication circuit" apply to the Raspberry Pi etc in the Caltech charger boxes? If so, should there be a partition between the power wiring and the Raspberry Pi stuff, making it two separate compartments?

I wonder what the NEC classification of the pilot wire circuit is (it shares the EV charging cable & connector with AC power, and it also connects to the LV side of the OpenEVSE board's isolation barrier.) Hmmm.

 

[FlasherZ]

Does "communication circuit" apply to the Raspberry Pi etc in the Caltech charger boxes? If so, should there be a partition between the power wiring and the Raspberry Pi stuff, making it two separate compartments?

No, there's probably an interpretation that would demand UL listing for its purpose, but it's legal to have certain small items in the panelboard. For instance, I have an X10 repeater mounted in my panel.

I wonder what the NEC classification of the pilot wire circuit is (it shares the EV charging cable & connector with AC power, and it also connects to the LV side of the OpenEVSE board's isolation barrier.) Hmmm.

The NEC applies to infrastructure wiring and doesn't apply inside the context of an appliance. However, it would likely be considered a class 2/3 circuit.

 

[doofenshmirtz]

I'm going to have to drive down from the bay area just to see thins installation.

 

[RubberToe]

Scouted this out today on bike. Very full, though some spots were ICE'd. Allowable per the signage, but some spots are strictly for EV's. The remaining half of the incomplete spots look to be worked on over the weekend as there are no parking signs posted starting midnight tonight.

RT

 

[Ulmo]

I believe that I have the math correct for calculating the actual output Voltage on the nominal 208V secondaries.

Odd. I can clearly see that you were right the first time, and I was wrong. :crying:

 

[MITE46]

Couple updates...here's a shot of inside one of the distribution panels...100A breakers

We started installing the stations on the adjacent wall...but this time we had to mount it to the columns, which made station mounting a little more difficult, but still possible.

Also, the ultrasonic sensors needed x-axis and y-axis adjustment because the stations are not centered in the space...

Easy enough for a toddler to use...

We've also been working on dashboards...here's the live status chart...we're combining ultrasonic sensor data with openevse status to determine which spots are available, charging, finished, blocked(ICEd), error, etc. We have a couple of stations in the unknown status, waiting for firmware updates on those.

Here is the per station power graph...the idea here is that our adaptive algorithm will change the amps_setting(yellow) limits per station.

Then a couple of gauges for transformers capacity...LOL not sure we need that many sig figs

Temperature min, avg, max in the last 3 hours vs max power over that time period...stations 18-22 are still on old firmware so temps are wrong...until I update.

Cumulative stats...energy delivered, miles delivered, Carbon avoided etc.

Per minute stats for the last 30 hours...just to compare to the previous day usage

Per hour totals...compares the whole week etc...Most people don't need more than 10kwh per session, and most sessions only last 2 hours. I think we could easily install 200x stations in this garage with 300kva transformer + scheduling...

More coming soon!

 

[FlasherZ]

Love it - thanks for the update!

 

[ggnykk]

Great work!! Just curious, why put all 50 stalls in one single location, but not 10 stalls in 5 locations etc? It would be great if this is spread all over the Southern California area.

 

[RubberToe]

Stopped by this weekend after hitting 98% at the local free municipal lot. Wanted to top off and just check the site out. The cool thing is the charger shows real time energy delivered and updates every second. I took about 500wh in 5 minutes.

I can answer the question about why all in that spot. Caltech provided a grant for the system installation, so the system was installed at the parking garage site.

I have heard that some utilities are pushing to get workplace and multi-dwelling systems installed:
Charge Ready: A Plan for California | Edison International

Whether SCE is doing this solo or whether they would consider anything like what the Caltech guys are doing is a whole nother question. Maybe the Caltech guys commercialize this, and if it ends up being cheaper than what others are doing someone notices and a business is born?

No matter what happens, you have to love seeing something like this out in the wild.

RT

P.S. Somewhat related... When I had my Level 2 charger installed in my condo garage, the city needed to do an inspection of the large transformer sitting in a room off our underground garage. No one ever goes in that room. There are 50 units in the building. I guess the idea is that because an EV could draw so much power, they need to make sure the transformer isn't overloaded. So I would imaging any large installation in a condo building will also have to take that into account. Whether the city will pop in a larger unit if required, and/or who pays for it is another issue. Or maybe the Caltech system insures that the highest power draw is within the limits of the transformer during all times?

 

[MITE46]

SCE Charge Ready is on our radar...we hope to participate as a provider soon =)

Couple quick updates...since this is our first deployment, we're going to add a couple of PowerLogic PM870 meters on both the 480v and 208v sides.

In the electrical room we have 4x PM870's, one for the building main, one for fan loads, one for each of the 150kva transformers. Each meter has an ethernet card so we can query it with SNMP, custom webpages etc. We haven't hooked up the 480v mains yet, so those are showing 0v right now, CTs are in though.

Here are the meters on the 208v side. We had to put in a communications box since the meters are quite far from the electrical room.

Meter connections, with CTs

Some 208v stats:

It tracks voltages, currents, power(real, reactive, apparent), power factor...we're still trying to make sense if what we are seeing is normal.

Network card...I was hoping to use modbus to query the additional units over SNMP, but that's actually not possible. For simplicity, we just installed separate network cards for each meter.

Also added another 8x adaptive stations last week...total up to 48x online now.

 

[chuck-h]

I see negative power readings...maybe CT's are installed with reverse polarity?

Chuck

SCE Charge Ready is on our radar...we hope to participate as a provider soon =)

Couple quick updates...since this is our first deployment, we're going to add a couple of PowerLogic PM870 meters on both the 480v and 208v sides.



It tracks voltages, currents, power(real, reactive, apparent), power factor...we're still trying to make sense if what we are seeing is normal.

 

[MITE46]

Yes most likely I have the CT polarity wrong...will have to address that on the next trip there.

 

[chuck-h]

SCE Charge Ready is on our radar...we hope to participate as a provider soon =)

Couple quick updates...since this is our first deployment, we're going to add a couple of PowerLogic PM870 meters on both the 480v and 208v sides.

In the electrical room we have 4x PM870's, one for the building main, one for fan loads, one for each of the 150kva transformers. Each meter has an ethernet card so we can query it with SNMP, custom webpages etc. We haven't hooked up the 480v mains yet, so those are showing 0v right now, CTs are in though.

With mains/feeder metering, you could implement the algorithm I was talking about earlier .. that is, regulate the chargers based on total feeder current, even when non-EV loads (e.g. ventilation) come on autonomously. This is a feature I want to see in my multi-unit residential condo installation soon.

 

[MITE46]

yep that was the plan all along...that's why we have meters on the feeder as well as circulation fans.

 

[chuck-h]

...
Network card...I was hoping to use modbus to query the additional units over SNMP, but that's actually not possible. For simplicity, we just installed separate network cards for each meter.

My ethernet interface retrofit uses character recognition!

 

[brkaus]

My ethernet interface retrofit uses character recognition!

View attachment 169763

I have to say this Caltech system is just so cool. This is really what parking garages need in cities, office parking garages, large hotels. One size fits all isn't going to cut it as there are more EVs charging. Putting in a large number of 80 amp chargers will be capacity constrained, lower power won't meet the needs for quick turnaround folks.

Love the idea of parking spots for anyone.

As for the photo above, hopefully in a future update, you can read the sub-meter directly, and not have to write it on post-it notes for the camera to read (sorry, couldn't help it).

 

[Elon'sCousin]

MIte46, I have to say you did a brilliant job designing and implementing this charging setup. I noticed the new 32A Clipper Creek chargers that were put in on Sunday. Nice! Is there any chance that we might get one or two HPWC chargers for Teslas?

 

[MITE46]

Yep, there were 8x HCS-40's installed last week as part of a new experiment. They are also load balanced with the scheduling algorithm.

This is a 8 station group shared on 100A circuit...still 1 plug per parking spot.

We do have a bunch of HPWC v2's we used for testing, but we probably won't install them until the end of the year. So for now you will have to carry your J1772 adapter. =)

We recently reached 20MWh delivered over 3.5 months which at $0.15/kwh is about $3,000 or $30/day or < $1/car/day

I fixed one of the CTs that was flipped in early May, that's why the reactive power is all weird.

Total station count is 53...will probably add more in the near future =)

 

[hcsharp]

Yep, there were 8x HCS-40's installed last week as part of a new experiment. They are also load balanced with the scheduling algorithm.

This is a 8 station group shared on 100A circuit...still 1 plug per parking spot.
...

How are the HCSs load balanced? I didn't think they had that feature from ClipperCreek. Are you measuring the load on each station and intercepting/overriding the pilot signal on each one? It looks like each one has a wireless antenna for your network. Why didn't you just use 8 more of your own stations?

 

[MITE46]

Yes, load is measured on each station every 5 seconds and the state is transmitted over Zigbee mesh to a local "load management controller". The measurements are processed along with other constraints like transformer/infrastructure limits and solved by a linear program. If adjustments are needed, a message is sent to adjust the pilot signal on that particular station. If communications fail for some reason, the failsafe mode kicks in which stations are automatically lowered to protect the infrastructure.

There's a couple of near term projects that require a lot of stations(>50+ each) and UL listing. We don't have the man power to build hundreds of stations right now. Nor to wait 3 months for UL approval. Our technology is in the algorithm and scheduling anyway...