Economic Benefits of Smart Parking Lots
Jurica Babic*, Arthur Carvalho+, Wolfgang Ketter+ , Vedran Podobnik*
*University of Zagreb, Croatia
+Erasmus University Rotterdam, Netherlands
Abstract The ever-increasing number of electric vehicles (EV) on the road promotes the idea of sustainable
transportation by reducing CO2 emissions. This inherently means there is a growing need for
charging stations as well. A potential solution to address the need for charging stations is to
transform traditional parking lots into smart parking lots, in a sense that smart parking lots provide
not only parking services, but also the possibility for EV owners to charge and discharge their cars for
a price. Due to the inherently complex and dynamic environment, a potential obstacle, from a
business perspective, to the process of transforming parking lots into smart parking lots is the
complexity of estimating the profit of the smart parking lot's owner and, consequently, the length of
time required to recover the cost of the initial investment. We propose a simulation-based approach
to estimate the smart parking lot owner's profit during a certain period of time. We use real-life data
from existing parking lots, charging stations and wholesale electricity market for analysing a variety
of different investment strategies. Using a set of the most relevant "what-if" scenarios, we discuss
the potential impact of smart parking lots to energy business.
Introduction and Background
The alteration of the vertically integrated markets into open, free and smart markets (Bichler, Gupta,
& Ketter, 2010) forms the energy landscape of the future. Electric vehicles (EV) will play a critical role
in the future energy landscape, which is well recognized by both US government (i.e., a target goal of
1 million EVs by 2015) and German government (i.e., a target goal of 1 million EVs by 2020). As a
consequence of increasing number of EVs on the road, there is a growing need for charging stations
as well.
A potential solution to address the need for charging stations is to transform traditional parking lots
into smart parking lots, in a sense that smart parking lots provide not only parking services, but also
the possibility for electric vehicle owners to charge and discharge their cars for a price.
With this perspective, the parking lot’s smartness comes from the extensive use of energy
informatics (Watson, Boudreau, & Chen, 2013) and thus behaves as an electricity retailer by acting as
a player on a target electricity market. A single EV is, to a certain degree, a prosumer (Khalen, Ketter,
& van Dalen, 2014), in a sense that it can procure electricity from its battery as well as consume
electricity. A single EV is not able to participate in the target electricity market on its own due to the
fact that it only has a modest amount of electricity available to buy or sell. However, multiple EVs can
form a virtual power plant (VPP) (Kumagai, 2012), and together act as a competitive player on the
target electricity market. Our proposed model tackles this issue by putting the parking lot owner in
the role of an electricity broker which trades electricity between EVs and the target electricity
market.
Due to the inherently complex and dynamic environment, a potential obstacle, from a business
perspective, to the process of transforming parking lots into smart parking lots is the complexity of
estimating the profit of the parking lot's owner and, consequently, the length of time required to
recover the cost of the initial investment. For example, the parking lot’s owner must deal with the
uncertainties related to the electric vehicles, including:
How many cars will be parked on the parking lot and for how long will they stay parked?
How much electricity a car owner is willing to buy/sell?
How much money a car owner is willing to pay/receive for a certain amount of electricity?
Given the current market conditions, what is the expected market price for a certain amount
of electricity?
Approach
We propose a simulation-based approach (Ketter, Collins, & Reddy, 2013) to estimate the parking lot
owner’s profit during a certain period of time. For ease of exposition, we define the smart parking lot
setting using three entities: (i) the parking lot owner (broker), (ii) the electric cars (customers), and
(iii) the parking lot itself (the physical structure). Since the smart parking lot itself is conceived as IS-
based broker (Peters, Ketter, & Collins, 2013), we define its innovative IS artifacts: (i) smart parking
lot model with entities and interactions among them, (ii) methods for smart parking lot management
regarding electricity trading in the context of smart markets, and (iii) operationalization of the
parking lot model in the form of simulation.
Figure 1 presents the smart parking lot ecosystem which consists of 3 entities, namely the Smart
Parking Lot, the Electric Vehicles, and the Electricity Market, and 2 relationships, namely the Smart
Parking Lot - Electric Vehicles relationship and the Smart Parking Lot - Electricity Market relationship.
We model entities and relationships through, respectively, agents and markets. The Smart Parking
Lot acts as a broker connecting both markets, as we detail later.
Figure 1: Smart parking lot ecosystem
The smart parking lot (SPL) is defined in terms of a variable representing the amount of money it is
willing to pay/receive for a unit of electricity at a certain timeslot. Clearly, this reservation price is
dependent on the current market price in the target electricity market. In summary, the SPL
performs the following hourly activities (Figure 2):
Calculation of free parking spots and queue size. If EV wants to enter the SPL, the SPL will
provide the EV with a free parking spot from its pool of free parking spots (if the pool is not
empty) or the EV will be put in the SPL.
Calculation of electricity price. In the beginning of every time-slot (hour), the SPL fetches the
current electricity price from the Electricity Market agent (EM) and uses its profit margin to
calculate its selling and buying electricity prices.
Payment for parking and (dis)charging services. The EV needs to pay to the SPL agent for
both parking and the electricity service provided.
Figure 2: Smart parking lot activities
Each electric vehicle (EV) is described by two variables: (i) the amount of electricity required by the
car owner when arriving at the parking lot at a certain timeslot; and (ii) the price the car owner is
willing to pay/receive for a unit of electricity at a certain timeslot. Similar to the smart parking lot,
the EV reservation price is defined in terms of the market price plus a premium representing the car
owner’s willingness to pay for using the smart parking lot facilities. The EV lifecycle is described with
a flowchart on Figure 3:
Figure 3: Electric vehicle activities
Calculation of parking duration. We model arrivals and staying at the parking lot using a
M/M/c/0 queue with time varying parameters, allowing different timeslots (hours) to have
different arrival rates and service rates.
Calculation of the amount of electricity an EV is willing to (dis)charge. We assume that the
amount of electricity an EV is willing to (dis)charge follows a normal distribution with mean
equal to 15 kWh, truncated at 30 kWh, the standard deviation being equal to 10. A positive
value of the (dis)charge quantity means that EV is willing to charge its battery at the SPL,
whereas a negative value means that it is willing to discharge, i.e., sell electricity. By setting
the mean value to 15 kWh, we mimic the real-world situation where more cars want to
charge their batteries rather than discharge.
Determining whether the EV is willing to (dis)charge regardless of price. This activity
decides whether EV will take into account the electricity price when deciding whether to
engage in (dis)charging. The electric vehicle EV will engage in charging or discharging,
regardless of the current electricity price with the specified probabilities. This way we mimic
the real-world situation where a car arrives at a charging station and needs to charge its
battery regardless of the price, e.g., because the battery is almost empty or because there is
no other charging station nearby.
Calculation of the reserve price for (dis)charging. In case of charging, the EV decides to
proceed with the transaction only if the EV's reserve price is higher than the current
electricity price offered by the SPL. In case of discharging, the EV decides to proceed with the
transaction only if the EV's reserve price is lower than the current electricity price offered by
the SPL. For the calculation of EV's reservation prices, we assume that EV has the alternative
choice of (dis)charging at home, where its home supplier forms an electricity price
analogously to the SPL, but with different profit margin. We assume that the profit margin
follows a normal distribution with mean equal to 0.2, standard deviation equal to 0.1, and
truncated at [0, 1]. Further, we assume that EV was parked at home for some hours before
entering the SPL in case of discharging or, in case of charging, that EV will be parked at home
for some hours after leaving the SPL. The exact amount of hours is drawn from a uniform
distribution with range [1, 12].
Determining whether the EV will stay parked longer to fully complete the electricity
service. This activity is event-based and triggered after an EV enters the SPL and decides to
(dis)charge a certain amount of electricity. If the electricity amount cannot be processed
under the given parking time and charger speed this means that there is enough time for the
EV to fully (dis)charge its battery during its parking. In this situation the EV has two options: i)
to partially (dis)charge its battery; or ii) to prolong its parking time until the full (dis)charging
is complete. The EV will go for the latter option of parking time prolongation with the
probability defined in the simulation setup.
Finally, for the parking lot itself, we need to take into account the number of available parking spots,
the number of arriving vehicles, and the amount of time each vehicle will be parked during a certain
timeslot. In order to handle the inherent uncertainty regarding these variables, we model the parking
lot using a M/M/c/0 queue with time varying parameters.
By allowing different timeslots to have different arriving rates and service (parking) times, traditional
equilibrium results in queueing theory are not always valid. Consequently, we have to rely on
simulations, instead of traditional queueing theory closed-form equations, to estimate the expected
number of parked cars and parking times throughout multiple timeslots.
Preliminary Findings
In order to simulate a real-life scenario, we derived arrival rates and service times from the work by
Ferreira et al (Ferreira et al., 2014). In particular, the timeslots in our simulations represent different
hours in a day and, consequently, we derived 24 arrival rates and service times. Also, we use real-
world electricity prices from the day-ahead Nord Pool Elspot market (i.e., prices from a year 2014).
Our analysis of the SPL ecosystem identifies potential consequences for the SPL business by
considering different investment pathways. In particular, we scrutinize the 9 different SPLs from the
perspectives of both electricity trading and extended parking due to the provision of electricity
service. Furthermore, we define the SPL utilization, which is an important key performance indicator
(KPI) that provides insights on the usage of the parking and electricity services.
Table 1: Scenario-dependent parameter values and results
It can be noticed that profits from electricity trading increase with parking size and charger speed.
Although this outcome is intuitive and somewhat expected, the low absolute values, including low
profit discrepancies between the 9 scenarios, might make one question about the profitability of the
SPL business. However, the overall benefit from the energy service is not only measured in terms of
electricity trading profits, but also from the extra time an EV was parked in order to fully complete
the (dis)charging operation.
Interestingly, the results show that the most profitable investment option is to buy the slowest type
of charger. The rationale behind this result is that, in comparison to other charger types, the slowest
charger increases the chance that the requested amount of electricity cannot be transferred
between the EV and the SPL within the initial parking duration. Another interesting point is that the
results show that the SPL's main source of income is due to the extended parking service.
We introduce three types of KPIs that explain how well a particular SPL is utilized. Parking utilization
measures how many EVs were parked at the SPL. Charger utilization is defined as the ratio between
the amount of electricity EVs (dis)charged and the maximum amount of electricity that could be
(dis)charged in case chargers from occupied parking spots ran at 100% rate while EVs were parked.
Electricity utilization is defined as the ratio between the amount of electricity EVs (dis)charged and
the potential amount of electricity that could be (dis)charged in case all chargers ran at 100% rate all
the time. Figure 4 shows the hourly mean values for the three KPIs in each scenario.
Figure 4: Hourly mean values for charger, parking and electricity utilizations
Current status of manuscript
We are currently fine-tuning our model, designing more comprehensive EV model and performing
more sensitivity analyses.
References
Bichler, M., Gupta, A., & Ketter, W. (2010). Designing Smart Markets. Information Systems Research, 21(4), 688–699. doi:10.1287/isre.1100.0316
Ferreira, M., Damas, L., Conceicao, H., D’Orey, P. M., Fernandes, R., Steenkiste, P., & Gomes, P. (2014). Self-automated parking lots for autonomous vehicles based on vehicular ad hoc networking. In 2014 IEEE Intelligent Vehicles Symposium Proceedings (pp. 472–479). IEEE. doi:10.1109/IVS.2014.6856561
Ketter, W., Collins, J., & Reddy, P. (2013). Power TAC: A Competitive Economic Simulation Of The Smart Grid. Energy Economics, 39, 262–270. doi:10.1016/j.eneco.2013.04.015
Khalen, M., Ketter, W., & van Dalen, J. (2014). Balancing with Electric Vehicles: a Profitable Business Model. In Proceedings of the European Conference on Information Systems (ECIS) 2014 (pp. 1–15). Tel Aviv, Israel.
Kumagai, J. (2012). Virtual Power Plants, Real Power. IEEE Spectrum, 49(3), 13–14.
Peters, M., Ketter, W., & Collins, J. (2013). Design by Competitive Benchmarking : Tackling the Smart Grid Challenge with Innovative IS Artifacts. In Conference on Information Systems and Technology 2013. Minneapolis, Minnesota.
Watson, R. T., Boudreau, M.-C., & Chen, A. J. (2013). Information Systems and Environmentally Sustainable Development: Energy Informatics and new directions for the IS community. MIS Quarterly, 34(1), 23–38.
Top Related