Andy Shatney, Energy Project Coordinator with the City of Santa Cruz, and James Lonergan, Project Engineer with Enovity, discuss energy data analytics and commissioning.
The following is a lightly edited transcript of this recorded conversation.
Gridium | Hello everyone. Welcome to our conversation with Andy Shatney, energy project coordinator for the City of Santa Cruz, and James Lonergan, project engineer for Enovity’s energy services division. My name is Millen and I’m with Gridium. We’ll be discussing the energy use and energy cost savings that result when utility meter data, building controls level data, and proactive, smart building operators are united behind advanced building operations.
Okay, for quick introductions, let’s hear from Andy and James. Andy would you like to lead us off?
“We just don’t give up, which is, I think kind of the key to retro commissioning in general. You can’t give up on it. You have to keep editing, basically. It’s like writing a paper, you have to keep reading it and editing it and making it better.” — Andy, City of Santa Cruz
—
Andy, City of Santa Cruz | Sure. My name is Andy Shatney, I’m the energy project coordinator for the City of Santa Cruz. I became involved in this project when my local PG&E rep contacted me about the smart program, which is a program that Enovity manages.
The initial analytics were done by Gridium and they identified three specific locations at the City of Santa Cruz. The police station, city hall, and our central library. These are the same buildings that I identified looking for potential candidates looking for retro commissioning. I wasn’t able to look at the library because it’s maintained by a different department, but ultimately that building actually went through the process and the city hall building was undergoing mechanical upgrades, so we chose to look at the police station.
—
James, Enovity | I’m James, I work for Enovity, for the energy services division. Enovity: we’re an engineering firm based out of San Francisco. We operate, maintain, and optimize facilities to create higher performance places, and we do this through energy services commissioning and facility engineering. I’ve been implementing our smart program, which is a PG&E funded third party program, for a couple years. And I was happy to work with Andy and the City of Santa Cruz.
—
Gridium | That’s great, and of course we’re looking forward to hearing you guys discuss the work that you have done together. And speaking of which, we’ll set the stage here.
At first we’ll discuss how energy data analytics can power advanced building operations. Then we’ll dive into an example, Andy and James will share their detailed insights on a 40 ton AC unit and the changes made at the police station and the energy use and energy cost reductions achieved there. And then we’ll wrap with key learnings.
So James, can you share a little bit about how data driven engineering services can help building operations?
—
James, Enovity | Sure. So one of the tools we use for our projects for our customers is smart meter demand data and associated analytics, and we do that in two ways.
The larger way is taking an entire portfolio of buildings for a customer for the City of Santa Cruz that could be all of their municipal buildings, and assessing those in terms of ranking and how well they perform compared to one another and how well they perform compared to benchmarks for similar buildings.
Then the other way that we use the analytics is once we’ve found buildings that we want to target for improvement, we can dig into the details of their use and potentially find the causes of energy excess use or at least find the time periods when excess use is happening and give us some guidance as to how we might improve that building’s performance. Sometimes with the analytics you can even see exactly where the problem is coming from, if it’s a scheduling issue or if you’re expecting to see, say, an economizer operating and instead you see big demand spikes during hot periods or during not hot, temperate periods from compressor operation.
Or sometimes we’ve found really high night time base loads from equipment that’s been overridden on hand for a long time and all that you can see just from the meter level.
—
Gridium | Great, and as I understand it, a certain component here involves prioritization. What role does that play?
—
James, Enovity | So with the City of Santa Cruz, or any customer really, we have a lot of meter data. We have meter data for a bunch of buildings.
Some of those meters are for processes like the wastewater treatment facility or a meter that does streetlights, and prioritization entails going through their portfolio, finding their highest users, finding the highest users that are poor performers vs. high energy users that are… such as the process uses, which really there isn’t necessarily as much opportunity for sequence revision or energy reduction. And then ranking them based on both overall magnitude of the energy use and the magnitude of the excessive energy use.
And we do that using a couple different analytic approaches.
One is to look at the daily average scheduling at the building. So when the demand ramps up and when the demand ramps down, and to compare that to the load, which you can also see at the meter level sometimes. Big spikes that drop off right in the morning when the equipment starts off can indicate the equipment starting up well before the people are there. Long tapering evening curves, those show poor shutdown scheduling.
We look at the temperature demand response, so as it gets hotter outside, how quickly does the building’s demand ramp up? At what temperature does the does it start to ramp up? We see a lot of buildings where it appears that AC’s coming on at 50 or 55 degree outside temperatures.
So those kinds of things help us prioritize which buildings are going to be the best candidates.
—
Gridium | Andy that actually makes me think, bringing up to a point you had discussed outside of this conversation was that you had identified some buildings yourself or some retro commissioning analyses, and then that united quite nicely with what James discovered in Enovity’s prioritization of the meter level data.
Andy can you share a little bit about how you identified those buildings and then how you guys settled on digging into the police station?
—
Andy, City of Santa Cruz | Well I used a similar process too, so we have an energy star portfolio, which is essentially giving us benchmark data, but when you drill in on the data with the smart meter. I was a graduate of the year long machine class through PG&E, so a lot of stuff that James is describing, I did the same thing but in a different way.
So we did a regressive analysis where we put temperature over the kWh and got the reciprocal so we could see essentially, as the instructors describe, a cloud version of when things are ramping up and ramping down in trying to identify if the equipment schedules coincide with the building being occupied. So the building that we specifically looked at, the police station, two things were really interesting to me from the higher level.
What the Gridium software was able to do, which I think is pretty amazing, is it identified these three locations, but of those two locations, there was solar on them. So the solar, when it’s producing power in a day, actually will spin a… well it doesn’t spin a meter, because it’s smart, but people say this, but it will actually read negative kW, so it was able to see, even with that, that the building wasn’t performing well, and I’m able to see that too because the base loads at night don’t actually really change.
There’s always a certain level load that’s going on all the time, regardless of what the scheduling or the occupancy of the building is doing and although the site that we picked, the police station, is a 24/7 building, basically everyone goes home at night.
So when Enovity came in, I had a lot of stuff that I had leftover from my commissioning class that I wanted to do, but I couldn’t do because I didn’t have the skills or the funding to go get a mechanical engineer and James and company are all engineers, and they were able to look at the things that I had and other things that they discovered on their own going through the process and looking at how the equipment was actually operating, and they were able to follow the ASHRAE guidelines and do the things that needed to be done, but able to do it in a way that is going to meet the code and be safe and get that savings we’re looking for without damaging the equipment or making the building comfort… you know, making the building so it operates the way it actually should operate.
So it was a really stroke of luck for me, because it came at the very end of this class and I really wanted to be able to do these things, but I just didn’t have the bandwidth to do it, so, yeah I feel extremely fortunate for this program and the software.
—
Gridium | That’s awesome. Can we dig a little bit deeper into what happened at the police station? James, can you set this detailed case study up for us?
—
James, Enovity | Sure. So, like Andy said, the police station fell out of our prioritization as the best building to do first with the smart program because of the high energy use and because of the control system they had.
What we found when we went in was the building has a couple of large air handling units that are fairly new that were put on as part of a larger equipment retrofit project. Those units are equipped with standalone Johnson controllers and the site also has an Alerton front-end controlling zone-level system.
So our initial approach was to look at the Alerton system because that’s probably the easiest place to implement retro commissioning changes and we can trend them and see them and compare them to our Gridium savings.
So what we did was we reduced the zone air flows, we scheduled some zones that weren’t being occupied at night, and we implemented a dual max airflow program, which allows the cooling airflow to ramp down to a minimum during outside air ventilation and for it to ramp back up to a lower heating maximum from that ventilation as going a transition from cooling to heating, reducing overall flow, especially during outside air ventilation.
We were pretty pleased with the results of that, but talking to Andy and looking at the analytics, we felt that there was a lot more we could do, so we dug into the Johnson controls on those Johnson units, and we also found, thanks in part to, well, thanks entirely to Andy’s help and trending, a lot of the points on those units, that the units were leaking supply air and affecting the outside air temperature sensor, causing the compressor to short cycle.
So, kind of as a second step in our process, we did our best to optimize the standalone controls on the unit, we improved the economizer control sequence, which dipped from enthalpy to outside air, raised the outside air lockout. We improved the exhaust fan control – that was not coming on at all. We temporarily plugged the leaks and then subsequently fully plugged, and then we relocated the outside air sensors so that it wasn’t being affected by the supplier by the compressors coming on, which helped reduce compressor cycling and overall compressor operation in general.
—
Gridium | As we were talking about this earlier, Andy you shared some pretty remarkable screenshots of what the result of all this was, and I think that plugging the leak story is actually kind of interesting, too, so if you could share a little bit more detail about what happened there?.
—
Andy, City of Santa Cruz | So, when we were up on the roof looking at these two 40 ton units, James was essentially getting into the controls, the standalone control, and really just figuring it out.
Reading through this 100 page manual and figuring it out, because this controller is 15 years old, so it doesn’t have a lot of the things that people are doing now as far as outside air resets and it’s kind of just not what we need. So he’s trying to really get it to work the way it should work and one of his coworkers, Robert Rodriguez, is standing next to the unit, and he’s like “Man! There’s all this cold air blowing out of the unit,” and just like a ton of it, and we’re like “Well that’s not right.”
So we looked at the supply air cabinet and instead of the duct that goes up into the unit itself, it just wasn’t sealed correctly. When they craned up the unit, they set it on its adaptor curve and it just wasn’t right, it wasn’t sitting right. So essentially, all the units are sending conditioned air out into the outside. It was really inefficient, so we had Johnson controls came, and they used double-sided tape and, I’m trending 550 points at the buildings, so I was able to see the VFDs on both those units, just looking at nighttime, literally drop like five — five command speeds or 5% just with that one measure.
That was a really huge measure and later on, that sticky tape kind of gave out and the original contractor came back and they sealed up both those cabinets with sheet metal, totally sealed it up, and so I was seeing on the low side of it, I’m looking at this… James, do you want to talk about the graph a little bit?
—
James, Enovity | Sure. Millen, could we back up? We can do the before and then the after. Yeah. So in the base case, you can basically see the minimum speed was about 35 at night, 35%, and the VFD trend is the red line, and then the daytime speeds are pretty consistent every single day, regardless of temperature, which is the green line, outside air temperature at about 65 to 70% speed.
With the variable error system like they have at the police station, you would expect two things: that the minimum at night would be much lower than 35% and that the maximums would vary more based on outside air temperature. You would expect lower airflow on temperate days, which this trend is 60-65 degree high temps outside, so you would expect to see more variation in those.
And so, in the after case, you can see that we’re getting down to 14% minimums, or even lower in some cases, and our maximum never gets about 60%, which is great. And it’s more variable, much more variable, which we were hoping to see at the police station, because they don’t have a typical office schedule. Things change every day there. You can see plugging those air leaks and reducing those air flows made a big difference.
—
Andy, City of Santa Cruz | And putting more of the zones on a 12-hour 5 day-a-week schedule instead of being on a 24/7 day-a-week schedule. That helped a lot, too, because those zones are closing down and reducing the fan speed, so all of the measures that Enovity did contributed to the after and it’s a huge difference.
—
Gridium | Yeah, and so as we think about the effects of these measures, like the changes you’ve made to the units at the police station, what do you all do to ensure that it continues and persists? Should we dig in a little bit about the drift reporting that you guys look at?
—
James, Enovity | Let’s dig into drift reporting. So one of the best things about analytics is that you’re getting pretty much real-time, maybe at the most a week delay, demand data showing the effects of your energy measures, and one of the things that Gridium does is provide these drift analytics, which allow temperature normalized, or weather normalized week over week comparison of energy use so that you can compare your post-project energy use to your pre-project energy use in similar timeframes and see how you’re doing.
And that’s not the only tool we used to ensure a persistence, but it’s the principal one.
And it allows us to not only make sure that our measure is still saving energy, but also to make sure that no other operational changes are affecting the energy use in the facility that we’ve worked on.
And we like to provide these to customers for six months to a year after retro commissioning measures happen just to make sure that we’re working together to see these measures persist, which is crucial with commissioning. The big thing now is continuous commission. With all of the data that we have, we should be able to troubleshoot issues that come up quickly and maintain good building operations.
—
Andy, City of Santa Cruz | Yeah, exactly. That’s for me as the building operator, I get the drift report and I look at it and it’s really easy to visually see what’s going on. I can see, is there a decrease? Is there an increase? What time period in the day is the increase? Is it an operational change? Are we running something because it was a scheduled change? Is there a piece of equipment that maybe is not operating correctly that we didn’t identify that’s doing something that we don’t know?
So it’s just, for me, it’s a visual cue to go, “Oh yeah, this is working,” or, “What’s going on?” It’s really simple to understand if you’re actually maintaining that savings is, is it persistent? And if it’s not, why? And I get these reports, Enovity sends them to me pretty much monthly from one of the three people that are my contacts over there, and it’s a really great way to see if you’re saving energy, which was the point of this.
—
James, Enovity | And oftentimes with these drift reports, we can see exactly our measure’s effect, especially for things like scheduling measures, even economizer measures, you can see in the example.
You typically would expect a peak demand reduction during the daytime in economizer measures. Since our site has solar, it’s not during the daytime, but you can see peak demand reductions in the blue circle.
But for other measures or measures that have interactive effects, we’re more using the graphics to kind of assess the overall energy savings from a meter level rather than trying to pinpoint the effect in the demand profile. And with a site like the police department, it’s even more complicated because their solar system is so large, like Andy said, they get below zero.
They get net positive production during the daytime a lot of the time, so when measures like our fan speed reduction are slowing the fans down during the daytime, the police department is feeding more electricity back into the grid, and lowering their overall utility bill, but that doesn’t necessarily show up on PG&E’s demand data. So some of the measures you have to take into account, especially with sites that have complex demand profiles like the police station, but they may not be accounted for in the analytics, obviously.
—
Gridium | Right, and so as we think about what you all have learned from this experience, I think our audience might be interested to know some of the things you’ve taken away, sort of the highlights from working with the City of Santa Cruz, and James can you share what you think those highlights might be?
—
James, Enovity | Yeah. So for us, we start with Gridium, we start with the analytics, so it’s kind of a top-down approach, and with that approach, it’s really important to share your findings with the people on the ground at the facilities and get their opinion on the causes and whether or not those are mechanical or operationally constrained issues.
So if you get to the site and find out that your scheduling measure’s a no go because they need those spaces open at night, that’s no good. So it’s important to keep the people who are running those buildings informed on what you’re seeing at the meter level and why you think that might be, so that they can help you understand the best way to address the issue. So it’s kind of a two-stage approach where analytics is your first pass and feet on the ground, crucial second step.
—
Gridium | Andy what do you think the highlights would be from your experience working with James?
—
Andy, City of Santa Cruz | To really have an open mind and be blank in your assessment, meaning that some of the things you may think are the cause are not.
For instance, the leak, or the outside air sensor. Those little things can mess up the whole entire building without really being empirical about what you’re doing and really diagnosing the operation of everything, to a certain extent, obviously you can only go so far with it.
The things that you may think are the problem are not the problem, or the things that most people identify as the low-hanging fruit, if you change the economizers or scheduling, all those things are kind of things that anyone would look at, but that’s not the case with this building because two or three of the things that are probably… we don’t know for sure because we did a lot of measures at once, but with the outdoor air sensor being in the wrong place, that cabinet leaking cold air to the outside, if we had just looked at the typical things, we wouldn’t have identified those things.
So I think it’s really important to really look at, hey how does this work and why is it working that way and is it working the way in which it should?
—
James, Enovity | I would build on that outside air sensor thing. We had talked about it earlier. That was causing the compressors to cycle hundreds of times a day, and so even on top of the energy savings from improving that sensor’s response, it’s allowing us to prolong the life of the equipment and to improve its operation as well to identify those issues and correct them.
—
Andy, City of Santa Cruz | Well and to go a little bit further with that, it had been operating that way for a year. Those units were a year old, and on both of the two units that had those sensors in the wrong spot, the crankcase heaters and contactors had broke because they were being overused.
So I mean, you’re absolutely right.
Now that it’s been changed out, it’s coming on four or five times a day instead of upwards of over 100 times a day, which is ridiculous, and another point, too, is that sensor is actually placed where the manufacturer wanted it, where the engineers for that piece of equipment said to put it.
So we had, we were like pulling teeth. We had moved it and then we were told “no, you did the wrong thing, it needs to go back to where it was at,” and we tried to explain that that’s actually a really bad idea, but people were saying it needs to go back because that’s where the manufacturer said to put it when the Johnson guy came out and we showed him and explained to him and could show him exactly what was going on, he said “no, you did in fact put it in a way better spot than it was.”
So that’s kind of what I’m trying to say is sometimes even when you know something’s wrong, if someone’s already said that’s the way it operates, they don’t want to listen. They just shut their minds down, they’re just stuck on whatever it is they think is right and then you have to just keep coming at it because you have to stick to your guns. And we were able to do that and it’s a lot better, so, thank you.
—
Gridium | Yeah, that’s fascinating. So we will wrap this conversation up. If you have questions, please email energyservices@enovity.com or info@gridium.com and, you know, from Gridium’s point of view, I’d like to thank both Andy and James for participating in this conversation and then also to our listeners for their attention.
