A sustainability-based approach to resource allocation in the Smart - - PowerPoint PPT Presentation

a sustainability based approach to resource allocation in
SMART_READER_LITE
LIVE PREVIEW

A sustainability-based approach to resource allocation in the Smart - - PowerPoint PPT Presentation

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work A sustainability-based approach to resource allocation in the Smart Grid Siddharth Suryanarayanan Colorado State University March 17, 2015 Siddharth


slide-1
SLIDE 1

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

A sustainability-based approach to resource allocation in the Smart Grid

Siddharth Suryanarayanan

Colorado State University

March 17, 2015

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-2
SLIDE 2

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Outline

1 Smart Grid 2 Sustainability 3 CPS-based SGRA 4 Results 5 Sustainability aspects of DR 6 Future work

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-3
SLIDE 3

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Speaker background

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-4
SLIDE 4

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

The emerging Smart Grid

Source: US Dept. of Energy Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-5
SLIDE 5

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Functional charactersistics of a Smart Grid

Self-healing from power disturbance events Enabling active participation by consumers in demand response Operating resiliently against physical and cyber attack Providing power quality for 21st century needs Accommodating all generation and storage options Enabling new products, services, and markets Optimizing assets and operating efficiently

Source: US Dept. of Energy Office of Electricity & Energy Reliability Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-6
SLIDE 6

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Demand-Side Management (DSM)

May refer to utility sponsored programs aimed at increasing energy efficiency and water conservation or the management

  • f demand

Includes load management techniques Alteration of electrical energy consumption that offers incentives to consumers and utilities Advantages:

Increased efficiency Deferred capital costs for utilities Lowered emissions Prolonged availability of traditional resources.

Key enabler: changing rate structures in distribution systems

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-7
SLIDE 7

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Demand Response (DR)

Programs that provide incentives to consumers for deferring/curtailing demand during peak periods

DOE EERE Federal Energy Mgmt. Program

Triggered based on system reliability or market conditions

Courtesy: Dr. P. Bauleo, FCU

Example (HotShot Program) Load management program of Fort Collins Light and Power

Contacting customers (radio signal, email, pager, and websites) Cycling A/C and shifting water heater loads (residential) Automated BEMS & discretionary load shedding (C & I customers)

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-8
SLIDE 8

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Sustainability

Sustainability refers to the ability of systems to maintain, support, and endure

  • Acc. to the UN Brundtland Commission (1987), “sustainable

development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs.”[source: Wikipedia] Sustainability applies to a lot of domains including energy, agriculture, economics, infrastructure, and politics.

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-9
SLIDE 9

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Three pillars of sustainability

3 pillars of sustainability are social, environmental, and economic development People, Planet, and Profits Sustainable efforts must be quantified on the impact on the above ‘pillars’

Sources:Wikipedia; Clker.com; EPSfiles.net; Fouzi.net Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-10
SLIDE 10

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Three pillars of sustainability

Source:Wikipedia Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-11
SLIDE 11

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Cyber-physical systems (CPS)

Definition “Engineered systems built from and depending on synergy of computational and physical components” Contemporary breed of systems capable of interacting with humans, expanding system capabilities via computation, communication, and control “Coordinated, distributed, connected, robust, and responsive” Example Smart electric grid, smart transportation, smart buildings

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-12
SLIDE 12

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

State of the system

Residential electricity sales to grow 24% in next three decades Growth in transmission systems is stagnating Curtailing peak loads is imperative for reducing costs of electricity

“A 5% reduction in peak during the 2000–2001 CA energy crisis would have reduced the highest wholesale prices by 50%” (Thomas Weisel Partners Equity Research Report, 2007).

Distribution realm expected to burgeon with new and diverse grid technologies (Brown et al., 2010).

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-13
SLIDE 13

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Challenges

Most bulk electricity markets require a minimum power rating to bid into market

Makes it impossible for individual residential customers to participate in deregulated markets Allowing retail customer access to wholesale market prices may increase price-elasticity of demand and thus increase volatility in the market (Roozbehani et al., 2012).

Residential loads are not easily controlled and need to be composed of a large portfolio to provide a strategic DR product (CEC, 2014).

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-14
SLIDE 14

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Solution

Aggregator A proposed for-profit entity in a deregulated market structure interfacing a DR market (DRX) with a set of customers Functions:

1 intelligently coordinate customer appliances away from the

system peak time

2 enact a noticeable change on overall system 3 optimize performance for profit

Incentivizes customers by providing an alternative pricing

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-15
SLIDE 15

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Smart Grid Resource Allocation (SGRA)

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-16
SLIDE 16

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

SGRA: Enabling technologies

Retail electricity market must be fully deregulated ⇒ customers can choose between suppliers Control and communication infrastructure including requisite cyber-security measure must be fully developed and implemented

This will allow for seamless exchange of information and coordination of customer assets

Customer base must be willing to participate with proper incentives

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-17
SLIDE 17

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Customer incentive pricing (CIP)

A proposed pricing structure that the aggregator would offer the customers to allow rescheduling loads Instead of paying the local utility, customers pay the aggregator the CIP for electricity

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-18
SLIDE 18

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Information needed for determining CIP

Customer For each schedulable load i, customer provides:

1 δi, the runtime duration (in 15-minute intervals) 2 pi, the average power rating (in kW) 3 ti,start, the customer scheduled start time 4 (Ai,start, Ai,dur), a 2-tuple that represents the availability

window for load i determined by the availability window start time, Ai,start, and the availability window duration, Ai,dur

Assumption

Aggregator knows the exact time a load will run if it‘s not rescheduled by the aggregator (i.e., the start time is deterministic). Future work will void this assumption by going to stochastic SGRA.

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-19
SLIDE 19

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Information needed for determining CIP

Aggregator The aggregator possesses information on:

1 γ(i, λ, t), a binary function that represents whether the

customer will allow load i to be rescheduled to time t with CIP λ (γ = 1) or not (γ = 0)

2 s(t), the forecast spot market price of electricity in the bulk

electricity market (in cents/kWh)

3 r(t), the forecast price of electricity from the utility company

(in cents/kWh)

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-20
SLIDE 20

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Determining the CIP

Aggregator‘s position Aggregator must find the following:

1 L, the set of loads the aggregator is rescheduling 2 ti,resch, the rescheduled start time for load i 3 λ, the CIP vector containing 96-elements for a 24-hour period

Note: Let I be the total number of schedulable loads. |L| ≤ I because the aggregator has information about all I schedulable customer loads, but it does not necessarily have to reschedule all loads.

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-21
SLIDE 21

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Aggregator profit (P) has 3 components:

P = N + S − B (1) Total income recd. for selling negative load in bulk market N =

  • i∈L
  • γ(i, λ, ti,resch)

ti,start+δi−1

  • t=ti,start

s(t)pi 4

  • (2)

Exact payment recd. from N would depend on policy (FERC Order 745) Assumption: Aggregator is a well-behaved agent and does not manipulate market by misrepresenting the sum of negative load.

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-22
SLIDE 22

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Aggregator profit (P) has 3 components:

Total income recd. for selling electricity to customers S =

  • i∈L
  • γ(i, λ, ti,resch)

ti,resch+δi−1

  • t=ti,resch

λ(t)pi 4

  • (3)

Total cost paid to bulk market for buying electricity B =

  • i∈L
  • γ(i, λ, ti,resch)

ti,resch+δi−1

  • t=ti,resch

s(t)pi 4

  • (4)

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-23
SLIDE 23

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

SGRA optimization problem

max P (5) subject to Ai,start ≤ ti,resch ≤ Ai,start + Ai,dur ∀i ∈ L (6) and ti,resch ∈ Z ∀i ∈ L (7) c ∈ R ∀c ∈ λ (8)

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-24
SLIDE 24

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

SGRA example

Pricing data for Saturday 7/9/2011 Utility pricing data forecast from ComEd Residential Real-time Pricing (RRTP), r(t) Spot-market pricing data forecast from PJM, s(t)

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-25
SLIDE 25

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Modeling the customer

Our simulation study used 5555 customers with 56642 loads Modeling customer behavior is non-trivial

1 Using a newly proposed α-model (Hansen, Suryanarayanan, et

al.)

2 Using a POMDP-based approach

Takes into account the following:

1 Customer comfort 2 A threshold value of difference between CIP and RRTP for

participation (in the α-model)

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-26
SLIDE 26

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Customer loads

Two types of loads are considered: baseline and schedulable (smart) appliances Baseline load: thermal (air-conditioning, water heaters) and

  • ther non-schedulable loads

Smart appliances: 18 generic appliances with probabilistic data for each customer

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-27
SLIDE 27

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Results

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-28
SLIDE 28

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Results

Schedulable customer loads correspond to 11.2% of total energy used by 5555 customers Aggregator profit using forecast data for 7/9/11 was $813.92 Aggregator profit using actual data for 7/9/11 was $947.90 Customers profit was only $460.31 (forecast) and $794.93 (actual). This is because of objective function set to maximize aggregator‘s profit

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-29
SLIDE 29

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

3d load graph of asset-by-asset DR action

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-30
SLIDE 30

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Heat-maps of temporal information on aggregator profits

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-31
SLIDE 31

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Spatio-temporal demand response

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-32
SLIDE 32

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Sustainability aspects of DR

Societal benefit: Reduction in peak load due to SGRA-based DR offers a system-wide benefit of maintaining physical security of the electricity grid Environmental benefit: Reduction in capacity factor of peaking generators (typically running on “dirty” fuels such as diesel) Economic benefit: Customer savings

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-33
SLIDE 33

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Future work

Expand method to include stochastic input and more load models Quantify equilibrium when customers engage in maximizing individual profits Quantify spatio-temporal performance using HPC

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-34
SLIDE 34

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

Acknowledgments

Timothy Hansen, graduate student, ECE CSU

  • Profs. A. A. Maciejewski and H. J. Siegel, ECE CSU
  • Prof. R. Roche, Universit´

e de Technologie de Belfort-Montb´ eliard (UTBM), France Funding from NSF grant numbers CNS-0905399 and CCF-1302693 and UTBM

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid

slide-35
SLIDE 35

Smart Grid Sustainability CPS-based SGRA Results Sustainability aspects of DR Future work

In memoriam

Professor Robert B. France

  • Dept. of Computer Science

Colorado State University

Siddharth Suryanarayanan Colorado State University A sustainability-based approach to resource allocation in the Smart Grid