1

AN AGENT-BASED APPROACH TO VIRTUAL POWER PLANTS OF WIND POWER

GENERATORS AND ELECTRIC VEHICLES

Guided by,

Mrs. Deepa M.UAsst. Prof.,EEE

College of Engineering Perumon

Presented by,

Arjun AnilS7 EEE

Roll No. 8

CONTENTS

1. Why Virtual Power Plants of Wind Power Generators and Electric Vehicles?

2. What Is Virtual Power Plants?

3. Wind Power Generators and Electric Vehicles

4. Storage Payment and Day Ahead Optimization Scheme

5. Experimental Results

6. Comparison Between VPP And Normal Wind Power Generators

7. Conclusion

8. References

3

Why Virtual Power Plants ofWind Power Generators and Electric

Vehicles?

Wind power generation has received

considerable attention in recent years

Enables the wind generators to counter the unpredictability of wind power generation

Supply of energy to the grid can be controlled based on the demand of energy

Increases the profit of wind farms

4

What Is Virtual Power Plants?

A VPP is a group of multiple energy producers and energy storage providers

The objective of VPP is to sell electricity as an aggregate

Participants of VPP:

1. Wind Power Generators

2. Electric Vehicles

5

VPP Participants

VPP is composed of some wind energy producers and electric vehicles as a single entity

VPP helps in day ahead marketing

In day ahead market the power is generated, stored and traded on day k-1 to deliver it on day k by the supplier

The supplier must ensure energy balance between generation and consumption

If there is any imbalance the supplier must pay the imbalance penalties

6

Wind Power Generators and Electric Vehicles

oWind power generators generate electricity depending on the weather conditions

oElectric vehicles store the energy produced by the wind power generators

oThe generated energy can be supplied to grid in 2 ways depending on the demand

1)Directly to grid 2)From energy stored in batteries of

electric vehicles

7

o The power generated for a given day by a single wind turbine( j ) of a wind power generation site ( W ) be Pj(t)

o The day is divided into N time slots, the expected electricity during nth time slot is given by

o The sum of overall wind power generation site W gives us the generation vector defined as

Where

o The expected generation vector z is used for deciding day ahead bid

8

W

Electric Vehicles-Batteries on Wheels

The function of the electric vehicles is to store the power generated by the wind power generators Lithium ion batteries are used in electric vehicles By using electric vehicles as storage, the power can be traded in day-ahead spot market EVs are characterized by a storage profile which defines the amount of energy stored in each time slot its battery (sv)

9

v = { v1,v2,……………………vk } be a set of electric vehicles

For an EV v Є V , let sv be the storage profile vector for N time slots

Where sv(n) is the quantity of energy that an EV can store at timeslot n

Since the EVs willing to provide at most units of storage

therefore

10

Storage Payment Scheme For Electric Vehicles

• Storage is provided to the Evs in the form of charging entitlements rather than money

• When the Electric Vehicles are used for storage, some amount of charge is left behind as payment

• The amount of energy given away is measured as a proportion of the amount of storage used (σ)

• This payment scheme reduces the depth of discharge

• Thus it will help to overcome the reduction of battery life due to participation in VPP

11

Day Ahead Optimization Scheme

• To place the bid the leader has to compute the following 5 parameters that determines the supply schedule

i) Amount of energy supplied directly to grid (x)

ii) Amount of energy transferred to batteries (b)

iii) The energy transferred from batteries to grid (d)

iv) Amount of battery storage capacity needed (y)

v) Amount of energy transferred to EVs as payment (g)

• If η is the batteries overall conversion loss, it is necessary to store 1+η units of energy to actually deliver 1 unit of energy

12

• The objective of the VPP is solving the following optimization problems

Where,

σ = g(n)/y(n)

η – Energy lost when electricity flows from grid to battery and vice versa

pe - Wholesale price of electricity

z(n)-Day ahead estimated generation13

- Revenues raised by VPP from the electricity sold in the

market

• is the net energy stored in the EVs batteries at beginning

• By solving the optimization problem the day ahead bid w is given by

w= x + d

14

Case Study

Electric Vehicle Data

• The cost of participation of EV in VPP can be given as

Where

cb - Battery capital cost

DoD - Depth of discharge

Es (DoD) - Energy that EV store on behalf of

VPP

LET – Battery lifetime in kWh

L(DoD) – Battery lifetime in cycles15

Since

Therefore cEV becomes

• If Ef (DoD) is the energy that EV receives from the VPP then the EV profit function can be defined as

16

Experimental Results

The main focus of the experiment was to assess the profit of VPP when compared with wind farm without storage

The profit gain of the approach is given by

Where

- Realized profit that VPP obtains

- Profit raised by wind farm without storage

The profit of VPP mainly depends on σ ,

σ = Amount of energy given to EVs as payment

Amount of storage used17

18

We take 3 values of σ – 0.05,0.1 and 0.15 When σ = 0.05 , the storage is relatively cheap and hence it is used widely to maximize the profit As σ increases the storage becomes more expensive thus it is less utilized and profit gain tends to shrink for σ = 0.10 and σ = 0.15 Another research objective is to assess the amount of storage needed to maximize the profit (graph b)

19

As expected, the amount of storage used decreases as it becomes more expensive

If we consider the highest level of demand in terms of storage the VPP must have a storage capacity ranging from approximately 50MWh for σ = 0.05 to 19MWh for σ = 0.15

(1) σ =0.05

(1) (2) σ =0.1

(2) (3) σ =0.15

(3)

From the graph, we can see that EV offers maximum profit when the DoD is 0.4

Therefore a single EV is able to provide a storage of 0.4x30=12 kWh (30kWh is the maximum storage capacity of EV)

Thus a VPP would need from 1583 to 4166 EV to store 19 to 50 MWh

From the results it is found that when σ = 0.05 , storage is relatively cheap and EVs are widely used

Although the price paid to the EV is lower, they make small but frequent profits throughout the year leading to high annual profits

• When σ increases, the usage of storage is less profitable to the VPP as it is less frequently used and annual profit is also reduced

20

Comparison Between VPP And Normal Wind Power Generators

Normal Wind Power Generators VPP With Wind Power Generators And Electric Vehicles

Less reliable

Intermittent and are prone to large forecast errors Low profit Simple in design, construction and supplying

More reliable and can compete with other mature technologies of energy generationGenerated power is stored and can be supplied whenever needed Higher profitMore complex

21

Conclusion

This paper shows a method to make wind power generation more reliable by forming VPP

The profit can be maximized by optimizing the schedule of supply to the grid

Introduced a novel scheme of paying the EVs for their storage through supplying energy at no cost

22

References[1] A. L. Dimeas and N. D. Hatziargyriou, “Agent based control of virtual

power plants,” in Proc. Int. Conf. Intelligent Systems Applications to Power Systems (ISAP-2007), 2007, pp. 1–6.

[2] L. M. Costa, F. Bourry, J. Juban, and G. Kariniotakis, “Management of energy storage coordinated with wind power under electricity market conditions,” in Proc. 10th Int. Conf. Probabilistic Methods Applied to Power Systems (PMAPS-2008), 2008, pp. 1–8.

[3] G. Giebel, R. Brownsword, and G. Kariniotakis, “The state-of-the-art in short-term prediction of wind power: A literature overview,” Project ANEMOS D1.1, 2003.

[4] R. Piwko, D. Osborn, R. Gramlich, G. Jordan, D. Hawkins, and K. Porter, “Wind energy delivery issues,” IEEE Power & Energy Mag., vol. 3, no. 6, pp. 47–56, 2005.

[5] J. F. Manwell, J. G. McGowan, and A. L. Rogers, Wind Energy Explained: Theory, Design and Application. New York: Wiley, 2002.

23

24