Smart EV Charging in Parkades: A Data-Driven Approach for Building Managers

Moustafa YoussefBlog

One in four new car purchases in BC today is an electric vehicle (EV), and with the Lower Mainland’s high concentration of apartment and condo buildings, more parkades will inevitably need electrification. If you’ve ever seen a parkade full of EVs, you might assume that charging them requires a massive electrical service. Sure, a transmission-connected community centre or IKEA can support a dozen level 2 chargers and a couple level 3 fast chargers on demand, but many apartment and condo buildings are still supported by relatively small transformers and aging distribution grids. In this article, I’ll explain how residential parkades can still maintain EVs with limited electrical supplies and introduce a simulator we developed that helps building managers determine the optimal supply capacity for EV charging in residential parkades before committing to any costly electrical upgrades.

Why switch to EV

Switching to an EV can reduce energy costs by up to 80% while lowering your carbon footprint. EVs have lower operating costs, fewer moving parts, and no oil changes, making them more reliable and cost-effective in the long run. With expanding charging infrastructure and BC Hydro rebates, owning an EV is becoming more convenient and affordable than ever. Plus, driving electric helps cut air pollution and reduce dependence on fossil fuels, contributing to a cleaner, more sustainable future. However, charging remains a key barrier to EV adoption—43% of Canadians who don’t own an EV cite “lack of access to home charging” as a major obstacle. This highlights a crucial opportunity for building managers to improve resident satisfaction by enabling convenient EV charging solutions.

Long parking dwell times allows for trickle charging

EVs spend most of their time parked—especially overnight. With limited capacity but plenty of time, power sharing between chargers can ensure that each EV gets enough charge to meet daily commuting needs. The ideas is that on workdays, the EVs replenish the small amount of energy used during their daily commute with a trickle charge during the night. And by the end of the weekend, when they’ve been parked for even longer, they’ll have a full charge by Monday morning.

How energy management systems can help

When residents return home in the afternoon and evening, they begin cooking and using other appliances which causes the building’s power demand to peak. As a result, there needs to be an energy management system (EMS) that controls the chargers and ensures that they don’t overload any of the building’s electrical equipment including panels, switches and transformers. As shown in the example below, when the charging demand exceeds the maximum charging capacity, the available power is shared among the parked EVs. This isn’t the only approach—different EMS can adopt various charging strategies, such as prioritizing EVs with very low charge levels.

Neighbour’s Multi-Unit Residential Building Charging Simulator

We developed a simulator to help building managers and stratas determine the amount of charging power needed to keep their residents’ EVs adequately charged, before committing to costly electrical upgrades. Our simulator models tenant travel patterns and predicts EV state of charge (SOC) over time, providing data-driven insights for smarter charging infrastructure planning. It simulates the behaviour of each EV tenant as well as the building’s overall demand, accounting for unique factors such as each resident’s expected daily departure time, arrival time, and energy consumption.

The simulation treats these variables as random to reflect real-world behavior and simulated the role of the chargers’ energy management system. It checks building demand while allocating power to each EV. The simulation also updated each parked EV’s SOC before its departure from the parkade the next day. So for example, if an EV departed with an SOC of 80%, and used 30% during its daily commute, its arrival SOC would be 50%, and if it receives an overnight charge of 20%, its SOC the next day when it leaves is going to be 70%, and so on.

Below are simulation results for a parkade for a 16 storey building in New Westminster with approximately 100 units. The parkade is going to be home to 20 EVs and will have maximum charging demand equivalent to only 3 chargers: 3x32Ax208V or 20kW.

The graph above shows the building and EV demand over the course of a day. As you can see, there are two load spikes: one in the evening around 5 p.m., and another in the morning around 7 a.m. Residents start arriving around 1 p.m., plugging in their vehicles and beginning to charge. At the same time, the building’s electrical demand increases as more residents begin cooking and using their appliances. Charging power is curtailed slightly as the building’s demand reaches the supply capacity around 4:00 p.m. (150 kW), causing the total demand to flatten out, thanks to the EMS.

Below are key parameters used in modeling departure and arrival times, as well as daily energy needs. In this scenario, I assume that EVs consume an average of 25% of their state of charge (SOC) per day, or about 16.75 kWh—equivalent to a one-way commute of 40 km. This energy requirement can be refined using resident-specific data, as it heavily depends on location. For example, residents of an Abbotsford apartment building are likely to have higher energy needs than those in Burnaby if both groups typically commute to Greater Vancouver.

#Simulation parameters
initial_soc=0.75 #soc when the EVs first leave the garage
battery_capacity = 67  # kWh per EV
arrival_start_hour = 12 # Earliest arrival time  
arrival_end_hour = 24  # Latest arrival time  
departure_start_hour = 5  # Earliest departure time 
departure_end_hour = 11.5  # Latest departure time 
arrival_std = 0.25  # Standard deviation for arrival times
departure_std = 0.5  # Standard deviation for departure times
delta_on_return_mean = 0.25  # Mean delta SOC on return 
delta_on_return_std = 0.1  # Standard deviation for SOC on return 

And what about the EVs’ state of charge? The results were impressive. As shown in the graph opposite: by the end of the week and with the power of just three L2 charger, more than 50% of the time the 20 EVs were able to maintain a state of charge (SOC) higher than 75%. In less than 8% of instances of EVs leaving the parkade, they had less than 25% charge, of which about 5% had no charge at all, suggesting that some residents will have to be supplemented by charging elsewhere.

The table on the right is a general approximation approach to how much power is needed to maintain a certain number of EVs in the parkade with the parameters noted above. These parameters are refined or replaced with individual resident data if available, which will help to provide for a more optimal supply design.

Number of EVs parked in the parkade overnightRequired charging capacity in equivalent L2 (32A, 208V) chargers
1-5One charger
5-10Two chargers
10-20Three chargers

BC Hydro minimum charging performance guidelines

BC Hydro provides several guides for EV charging in residential parkades, including one on preparing for the EV Ready Plan Requirements. It includes a table (shown below) that squares the required supply capacity of a parkade with the annual distance travelled by the building’s tenants. These minimum charging performance guidelines are a great starting point for building managers and stratas approximating their future requirements. For example it suggests that a 100A breaker, which would have a capacity roughly equal to three chargers used in the simulation above, can support up to 31 EVs if they travel only 25 km per day.

Smart planning: simulation optimizes EV charging supply capacity

Planning to electrify a parkade for its residents’ existing or future EVs involves assessing its existing electrical service, building demand patterns, and then making recommendations for equipment and services suited to a building’s operating and administrative needs. Determining the right electrical supply capacity for EVs in the parkade isn’t just about avoiding unnecessary upgrade costs—it’s also key to minimizing overall electrical expenses, including operating demand charges that large electrical customers pay for. If you’re interested in exploring EV charging options for your parkade, let’s have a conversation.

Are solar designers picking tilt angles for solar panels properly?

Moustafa YoussefBlog, Video

Picking solar array tilt angle ground mount

Ground mount designs are more expensive and complex and require more engineering than flush mounted systems. Solar designers need to be extra careful when designing tilt angles for solar panels as there are several variables that if not tuned properly can result in significant space wastage and sub-optimal financial performance. One of the main variables in ground mount designs is tilt angle. In this video, I’ll create a geometric model that factors in both installation and real estate costs to determine the optimal tilt angle for solar panels, minimizing the system’s total cost.

How does a solar design properly pick a tilt angle vis-a-vis opportunity cost of the space that’s going to be occupied by the array


120 percent rule for solar breakers

Moustafa YoussefBlog

load_side_connection_CE_code_64_112

Most solar PV systems are connected to a property’s load panel or sub-panel. Solar inverters feed AC electricity to connected appliances, and any extra power is back-fed to the grid to power other appliances in the network. However, the Canadian Electrical Code limits the connecting capacity of multiple power sources feeding a distribution centre. In this article I will explain why such limitations exist and ways to increase a solar PV system’s capacity beyond CE Code 64-112’s 120 and 125 percent rules.

Why load side connections?

Load side connections are the simplest way to connect a solar PV system to an AC grid. The only upgrade required to a load panel is an installation of a bi-directional breaker as far away as possible on the bus from the other source breaker(s).  However 64-112 limits the sum rating of the supply breakers at 125% of bus rating for homes, and 120% for non-dwellings, and I’ll give an example to show why.

What’s the point of CE Code 64-112’s 120 and 125 percent rules?

Consider a bus that has a rating of 225A and connected to a single source of energy (transformer, generator, or what have you) through a 200A breaker. No problem here. The bus is always protected because if more than 200A of current tried to reach it, the supply breaker will trip and interrupt the current. Is this still true when the bus is connected to another power sources? Below is a picture of a load panel being fed from two sources, namely a solar inverter and another power source, which could for example be a generator, a transformer or another panel board. Each power source is connected through a breaker that limits the current or power flow reaching the bus.

solar inverter interconnection
This is how most small grid-tried solar PV systems (<500kW) are interconnected to the grid. They are installed behind a breaker like any other appliance connected to the bus.

Suppose that a solar inverter is connected through a 100A breaker. Now if both sources were feeding the bus at their maximal capacity, they could be pushing a total of 300A into the bus, which exceeds the bus rating of 225A :

The purpose of CE Code 64-112’s 120 and 125 percent rules are to protect distribution nodes from multiple sources of power.

For a 225A bus in a home,CE Code 64-112 requires that the sum of the breakers feeding a bus can’t exceed (225+25%) 281A as per CE Code 64-112 (4). So if the main breaker that’s connecting the home to the grid has a rating of 200A, the solar breaker(s) can’t exceed a rating of 81A or 80A. So we can install one 80A breaker to tie-in one inverter or for example, 4x20A breakers to tie-in four separate inverters or distribution panels connecting several inverters.

CE Code 64-112’s 120 and 125 percent rules

64-112 (4) limits the sum of the sources feeding a bus to 125% of the bus’ rating for homes, and 120% for non-dwellings. If we want to install a greater output capacity, we can

  1. buy a panel with a greater bus ampacity. This is the most expensive option, and its impact is limited. For example it’s not easy to find single phase panelboards with rating greater than 225A. We can also
  2. reduce the rating of the main 200A breaker. This depends on whether or not the client can afford to reduce their loading capacity.  Again this option is limited.

History of line vs load side connections in Canada

When Ontario spread-headed the adoption of distributed solar in 2009 with their microFIT program, they required output of solar inverters to be connected to a separate meter. A home would still have one run from a transformer, but it would split to two meters, one for consumption, the other for generation. Most jurisdictions today including Ontario are providing either net-metering or net-billing settlement systems that don’t require a separate meter. This load meter just has to be replaced with a bi-directional one to be able to measure exported energy.

Advantage of line side connections

Many installers and inspectors outside of Ontario are only familiar with load side connections because it is the most popular way of interconnecting relatively small systems. But CE Code 64-112 actually allows for line side connections (rule #1) before it allows for load side connections (rule #4). A line side connection will bypass any load-side connection limitation, including CE Code 64-112’s 120 and 125 percent rules.

CE_Code_64_112_Line_Side_Splitter_Solar_Grid_Tied
As we can see with a line side connection the panelboard is now fully protected and risks no overload. With a line side connection the panelboard no longer limits the interconnecting capacity of a grid-tied solar PV system. Furthermore, the line side equipment such as the meter socket doesn’t have to be upgraded at all, unless of course the capacity of the solar inverters exceeds the ampacity of the meter socket.

Line side connection on meter socket

If the bus bar on meter socket already has enough extra space for the new cables you want to install I would just install them there. It’s important to note to the inspector you are not actually increasing the capacity of power flow through the meter socket since solar is a negative load. So unlike a load side connection that increases the current through the bus, any current from the solar on line side will decrease the current from the grid and so the ampacity of the meter socket need not be greater than what it already is. Of course that’s assuming the size of the solar pv breaker is less than or equal to the rating of the main breaker on the home panel.