smart grid cstep

7/28/2019 Smart Grid Cstep

1/45

Smart Grids: Present and Future


NPTI, Bangalore4th July, 2011

Dr. Rahul Tongia, with Mohd. Saquib and H S Ramakrishna

Center for Study of Science, Technology and Policy (CSTEP)


2/45


Background

Basics of a Smart Grid Focus on what aspects are applicable to the

particular utility

Aside: most people worry about the what and thehow, ignoring the important question of why


3/45

3


Smart Grids

A smart grid delivers electricity fromsuppliers to consumers using digitaltechnology to save energy, reduce costand increase reliability.

Wikipedia

(More formal definitions arefarmore complex)


4/45

4


A Smart Grid is a Transformation of the power systembased on harnessing digital communications and

control Utilities will be able to:

Know what power is going where, and when

Charge appropriately for it

Control the use of (if not flow) of power

Although Advanced Metering Infrastructure (AMI)

is considered to be the basic building block for aSmart Grid, the Smart Grid is not just AMI!

The Smart Grid is a much broader set of

technologies and solutions


5/45

5


CentralizedGeneration

TransmissionNetwork

SupplierTransactions

Inter-Connections

DistributionNetwork

Meters &Displays

DistributedGeneration

ElectricVehicles

Loads andAppliances

EnergyEfficiency

Micro-Generation

ConsumerBehavior

CONSUMERDEMAND

Smart Grid

Smart Metering /AMI


6/45

6


India Smart Grid Forum (ISGF), which is a non-profitvoluntary consortium of public and private stakeholders,

was launched on 26th May 2010.

Also, India Smart Grid Task Force (SGTF) is formed,

which is an Inter-Ministerial Group and will serve as afocal point for activities related to the smart gridtechnology. Shri Sam Pitroda, Advisor to PM on PublicInformation Infrastructure & Innovation is the Chairman

for Task Force.


7/45

7


India Smart Grid Forum Working Groups:

WG - 1 - Advanced Transmission (incl. PMU, WAMS, FACTS etc.)

WG - 2 - Advanced Distribution (incl. SCADA / DMS, Distribution / Substation

automation, Power Electronics, FLISR, islanding, self healing, distributed

generation/renewables, etc)

WG - 3 Communications WG - 4 - Metering

WG - 5 Consumption and Load Control (Demand Response, Home Automation,

Appliances, Storage, Vehicles etc.)

WG- 6 - Policy and Regulations (incl. Tariffs, Finance etc.)

WG - 7 - Architecture and Design (Standards, Interoperability, Security, CIM etc.)


8/45

8


India Smart Grid Task Force Working Groups:

WG -1 - Focus on Trials/Pilots on New Technologies & Ideas

WG -2 - Focus on loss reduction and theft control including data gathering and

analytics, energy accounting

WG -3 - Focus on access of power to rural areas and reliability & quality of power to

urban areas

WG - 4 - Focus on distributed generation and renewable

WG - 5 - Focus on physical cyber security, standards & spectrum


9/45

9


Broad Aspects of Smart Grids

Generation Distributed Renewable

Transmission Improve transfer capacity Reliability (avoid blackouts)

Distribution {Includes consumption} Area of most effort

One aspect is smart metering Others include Demand Response aka Load Control Dynamic instead of mere DSM


10/45


Advanced Transmission

PMU, WAMS,

FACTS

Remote monitoring and operation of SubstationsStrategic Asset management

Self Healing Power Systems

Adaptive Islanding Systems

Advanced Distribution

Automation: FLISR, substation automation, SCADA/DMS

Integrating generation & Storage: Renewables, Distributed

Operation: Islanding (micro-grids),

Control: Power Electronics:STATCOMStrategic Asset management

Smart Grid integrates SCADA, AMR, GIS, ERP, Smart Substation Management System(SSMS), Advanced Metering Infrastructure, etc.


11/45

11


Status of transmission today

What are the exact Transmission losses today? (is it

known in every state?) How do these vary, and why?

Seasonality (loading)

Source of supply?

Is transmission congestion an issue?

New generation capacity

Increased loading

How can we price for congestion and impact ongrid?


12/45

12


What would happen to our grid if, say,

wind becomes 25% of the capacity? What are the options to deal with this?

More (fast ramp) supply

Curtail demand

Load sheddingSmart systems / Demand Response


13/45

13


Drivers for Smart Grids

US and Other

Developed Countries Meter reading Grid modernization

Robustness

Saving $$ Deregulation exposed a lot

of costs

Some consumers saw 20-40% increase in tariffs

Needs Time of Use (ToU) ifnot Real Time Pricing (RTP)

Indian

(Developing Country) Power system haschallenges

Loses Rs. 1+/kWh onaverage

Supply


14/45

14


Future (or even Subtle) Drivers US and Others

Carbon and green Bi-directional power

(Plug in) Hybridvehicles

New services

Home automation

Home monitoring

Green Power

India

Remove the humanelement in operations

The peak is NOT industrial

Smart peak management No more load shedding

Even in emergencies canallow smart control

LEAPFROG


15/45

15


What Smart Grids really mean

Cost Implications* ?

More choices

Includes renewables

Better quality and service

Greater resiliency / robustness

Increased efficiency and assetutilization


16/45

16


Fundamental Qs for the Regulator

Is a Smart Grid worthwhile? Cost Benefit Analysis

Who should pay for it?

High capital costs

What changes are needed in pricing models?

Variable if not Dynamic pricing

Need to reflect the peak *marginal* cost of power

To what extent must the solutions be deployed? Can theutility optimize based only on

Geography

Consumer, etc.? [80:20 rule]


17/45

17


Indian Examples of Functionalities

Loss reduction Requires precise and full metering

15 minute or 30 minute or even hourly

readings can help give visibility for operations Ending load shedding

Only two options

Buy more (peak) power Reduce Demand

(Third Option is to load shed!)


18/45

18


Drivers for Smart Grids Rhetorical Q: if developed nations dont have high AT&C

losses, and no load shedding, why do they need a smart

grid? A smart grid is about more than the above

Labor costs are an issue in the West Renewables and electric vehicles are high on the agenda in the

west, esp. Europe The regulator may not mandate smart grids

May only require smart meters May also require ToU tariffs or renewable integration

This de facto requires some level of a smart grid

Many nations have put in Smart Grid/Smart Metermandates (legislation), e.g., EISA (2007) in USA India does not yet have any legislative / policy support for smart

grids


19/45

19


Peak is growing faster than average(Independent System Operator-New England [ISO-NE] Example)

[Source: Kathleen Spees, CMU/CSTEP]


20/45

20


Peak Load in ISO-NE Change Between 1980 and 2006

[Source: Kathleen Spees, CMU/CSTEP]


21/45

21


21

Variability in Demand (NY)

Source: Walawalkar et.al 2007


22/45

22


0

1000

2000

3000

4000

5000

6000

7000

Load (MW)

Hours in Year 2008

Load Duration Curve - Karnataka

8760

6648

0

Load shed = 1150 MW


23/45

23


What is the Value of one kWh AVOIDED?

It could be from rooftop PV or smart grid or

anything Todays system for both CONSUMERS and

UTILITY are based on average cost accounting Ignoring cross-subsidies even

What we want is the marginal cost Costly power = UI, Power Exchange, IPPs, Diesel, etc.

The answer depends on when, where, etc.


24/45

24


0.64

1.321.77

6.30 30.57 1.92 10.72 0.00 33.040.03

1.121.40

0.810.70

4.16

0.07

0.11

12.251.84

0.76

16.73 0.16

1

2

3

4

5

6

7 8

9

10

11

1213

14

17 18

1920 21

22 23 24

2.025

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

Rs/Unit

Purchased Power (MU)

1. RAYALSEEMA(D)

2. DG PLANT (D)

3. TATA (IPP)

4. Non-Conventional

5. RTPS (C)

6.GERUSOPPA (H)

7. BTPS (C)

8. ALMATTI (H)

9. CGS (Mostly C)

(C) = Coal

(H) = Hydrp

(D) = Diesel

HYDRO(10-24)

10. TB

11. KADRA12. KODASALU

13. MGHEJOG

14. SHIVASAMUDRA

15. GHATAPRABHA

16. MUNIRABAD

17. VARAHI

18. MANI DAM

19. SHIMSA20. NAGJARI

21. SUPA

22. L INGANMAKI

23. SHARAVATI

24. BHADRA

Total: 126.4 Million Units (MUs)

Average Purchase Price: 2.025 Rs/Unir (Gross)

Power Purchase - KN - April 3, 2009


25/45

25


Buying Peaking Power

Peak power is always more expensive than theaverage Plants operate at only 500 or 1000 hours per year Ignores 15% target spinning reserves, today articulated

as 5% by GoI Blending such peak power today is what the Westdoes Raises the costs for ALL users for ALL kWh

KN example Raises purchase cost for utilities by Rs.1/kWh!

Alternative peaking tariff let those whocontribute to the peak pay for it Requires appropriate metering


26/45

26


Cost by utility

Source: Wartsila Report (2009): Real Cost of Power


27/45

27


A Smart Grid needs Smart Tariffs Short run: Pilot Long run: full-scale deployment

Today, limited off-take for ToU (voluntary, bulkconsumers) Differential appears too low to be attractive

Tariff Options Time of Use/Time of Day

Seasonal adjustments

Real-time Likely to be complex

Can allow selected RTP signaling like critical peak pricing (CPP) rare conditions

Can a utility undertake tariff innovations in a selectedarea or for selected consumers?


28/45

28


Some Hard Policy Qs Business case

If it made sense, wouldnt utilities already do it? The numbers depend on many unknowns (Time horizons, Consumer

responsiveness, Future tariffs and costs, Discount rates, etc.)

Incentives to participate Utility

If they are on a costs-plus regulated world, why do they care? Global experience has been capital-centric

Consumer Unless I am paid to modify my behaviour, why should I change? ToU or even real time pricing

Need much more than voluntary, small differentials

There are many challenges in policy Transfer of social welfare even if just a few people participate, EVERYONE can

benefit There will be some winners and some losers now what? How much should the schemes be mandatory vs. voluntary; opt-in vs.

opt-out? Privacy and Security

At the very least, the utility will know if a consumer is home or not


29/45

29


Costs and Benefits are Hard to Calculate

Investor (utility) Return on Investment is somewhat easier thansocietal impacts

Selected difficulties Long timespans Uncertainty of participation and effectiveness Cost allocation for Smart Grid vs. Grid Upgrade

Societal Cost-benefit is needed E.g., Improved power quality helps the consumer No need for diesel generator/inverter backups

Rigour is more than academic Confounding factors include annual load growth, seasonal

variations, unusual events, etc.


30/45

30


What do we need for a CBA?

Cost Benefit Analysis needs ALL costs (monetary, non-monetary, etc.) to ALL stakeholders across the life of the

project How do we convert implicit or value-laden impacts (e.g.,

time)? Assumptions

Challenges Different time periods

Different values by different people

High uncertainty (performance and more)


31/45

31


Framework for Cost-Benefit Analysis

Costs

Pilot costs are alwayshigher than in full-scaledeployment

Depend heavily on currentstatus of grid readiness

Benefits AT&C loss reduction

Freeing up capacity (peak)

Avoiding load shedding

Avoiding blackouts

Improved power quality

Load planning

Asset optimization

CRM benefits

etc.


32/45

32


Role of the Regulator Balance the needs of suppliers with consumers Assumption: Utility is to make a regulated (stipulated) return at best,

assuming performance targets (e.g., AT&C improvements) Any increase in tariffs (peak) must be balanced with a commensurate

decrease (off-peak) QUESTION: WHAT ABOUT TRANSACTION COSTS?

There are two types of tariffs wholesale (utility buys) and retail

(consumer pays) It is very problematic to allow one to be market while the other is purely

regulated (e.g., California crisis) Must have a plan in place for both

Suggestion: make both dynamic, reflective of the dynamic cost at

the margin (by time of day) Does the regulator want to cap consumer liabilities?

E.g., cap on peak rates (not allowing market full pass through)


33/45

33


Beware Parmenides Fallacy{Comparing the future to the present, instead of

alternative futures} Todays and Smart Grid future are not easily

comparable Latter may have no (feeder level) load shedding

A 15 minute automated reading cannot be comparedto todays monthly manual (often out-sourced)reading

Clearly, saving the Rs. 1-3/month for the meter reader is notsufficient to justify a Smart Grid/AMI

BUT, the AMI enables many new functionalities, such as Load profiling Energy audits / loss reduction Power purchase planning Outage detection, etc.


34/45

34


There are alwaystrade-offs:

Integrated Design

34

BusinessCase

Policy /Regulations

Technology

DESIGN


35/45

35


Choosing an Architecture Each utility should optimize based on factors like

Needs / goals Fuel mix

Consumer base

Legacy equipment, etc.

Considering just an AMR example, differentutilities have chosen RF mesh, PLC,GSM/GPRS, optical fibers, etc.

Adage from the IT world: Cheaper, Faster, Better Pick any two


36/45

36


Business Model Issues Smart Grids are Capital Expenditure (capex) heavy

Benefits accrue over time

Utility has 2 main choices (esp. given most are cash-strapped) Treat capex into the rate base for RoR calculations

Raises tariffs on paper Reduces rise in future tariffs due to monetization of benefits

Undertake outside funding Loan has debt servicing implications Grant (from state or central govt.)

Limited in availability, and unlikely beyond a pilot

Public-private partnerships (PPP)

Whats in it for a private player? Sharing benefits (ala ESCO model)

ESCO models

Require very strong calculations of baselines and metrics (targets) Baselines must be over 1 year long due to annual growth and seasonal variations

(forget if it is an election year!)

Irony the worse the present condition, the easier it is to justify a Smart Grid(e.g., loss reduction)

But one has to be honest in what is due to a Smart Grid vs. improved operations


37/45

37


The Future will be Different

Consumers may become generators

Possibly with distributed and/or renewables

IT has improved to the point where even smallerconsumers (residences) can meaningfullyparticipate in utility load control schemes

Residences are a major part of the Indian peak The future should have ZERO load-shedding (at a

feeder level, at least)

Discrimination acrossand withincategories ofconsumers is present today and may remain

Lets do it more intelligently

Incentivize behavior at the margin

Provide a minimum assured supply 24/7


38/45

38


Rethinking Quality

Today, consumers face load-shedding and numerousmomentary interruptions NOT captured in declaredKPIs like SAIDI, CAIFI, etc. Recommend adding MAIFI Recommend adding scheduled and un-scheduled load shedding

data, and making this public

A Smart Grid can immediately end feeder-level load-

shedding! Load limiting control switch integrated into meters (remote

controllable connect/disconnect) Quality impacts consumers

Diesel and backups Pumpset burnouts (est. implicit costs Rs. 0.50/kWh or higher) Can one split the benefits between utility and consumer? E.g.,

Normal tariff is, say, Rs. 5/unit, and diesel costs Rs. 14/unit Above a minimum assured supply, during shortage periods only,

charge a premium for unrestricted supply on a voluntary basis, e.g., Rs.

10/unit (or enough to cover the utility costs)


39/45

39


How to move ahead? Utilities must propose a roadmap/plan for smart

grids What functionalities are desired?

Why (use/business cases)?

What is the architecture and cost?

Pilot deployments Learning Pilots Learn about technology, its impact (benefits), consumer

participation rates/happiness, etc.

Deployment pilots Worry about price-points, integration, scalability, etc.

Since we dont know the best solution, we mustexperiment, learn, and iterate


40/45

40


Challenges in doing a Pilot

Pilot may be limited to off the shelf components/design

Need vendors and partners with appropriate experience and

expertise Design goals

Open standards

Scalability

Modularity

Must rethink the entire ecosystem of providers

This is not like R-APDRP There is no SRS or template

The solutions are evolving and must be iterative

Lowest Cost per se is a false choice Lifecycle costs matter

Performance (functionality) matters

Pilots will always be more expensive!


41/45

41


Pilot Projects:Possible Varying Functionality in stages

(not necessarily linear)

Smart Metering Reliability and Robustness (supply switching)

Renewables, storage, and distributed generation

Load control and Demand Response Smart Appliances

Signaling to consumers and devices [whocontrols is TBD]

Sensor networks, etc.

ICT for Power Systems:Accounting Auditing Monitoring Control

(R-APDRP)


42/45

42


Selected Qs for the DESIGN

What are the metrics, both prior and targets?

Input side and output side

What functionalities are to be deployed? How can we incentivize participation?

Specialized tariffs needs regulators approval Suggestion: supply and quality guarantees

How deep does the utility want to go?

All consumers?

Inside the home? - Direct control vs. economic incentives Suggestion: hybrid mechanisms

How can we gather the right data?

Granular data is missing before policies can be implemented Average cost of supply numbers need updating

Only dynamic systems can measure consumer contributions to the peak OR theirresponse to pricing

Information overload is a real challenge need good analytics if

not Decision Support Systems (DSS)


43/45

43


Services are the key to Transformation IT is a means to an end?

Same with power! Heat, Cool, light, move, etc.

Can we value not just megawatts but negawatts?

(else Rate of Return thinking limits value of savingpower)

Should we focus on making a car 10% or 30% more fuel

efficient or lower carbon emitting? What about getting rid of the car entirely?

Its not about a more efficient air condition per se what aboutredesigning for passive cooling?


44/45

44


Thinking of the FutureWe need Smart Grids

Business as usual (BAU) will not be sustainable

Adding supply is necessary but not sufficient must makeconsumption smarter

Consumers must see and behave based on not just their

average costs but their incremental impact on the grid This will create a few losers but (hopefully) morewinners

Appliances and consumption will become smarter

Whirlpool announced that by 2015 ALL their selectedhousehold appliances will be smart grid capable(worldwide)

Its not a question of when, not if


45/45


Questions?

smart grid cstep

Documents