The World Bank

1

Luiz Maurer AFTEG - April, 2006 lmaurer_at_worldban
k.org
2
This presentation has the following major
objectives

• Illustrate different types of shortages
• Comment on case studies and best practices
• Highlight the example of Brazil
• Stress the importance of demand management for a
  sensible power rationing
• Discuss opportunities to help other countries
• Present a forward approach (Chapter 7)

3
This presentation does not have as objectives

• Criticize how countries have been handling the
  challenges of power rationing (or Banks support)
• Imply that power rationing crisis
• Advocate that crisis management should overshadow
  the importance of long term adequacy and
  reliability planning
• Assume that the Brazilian solution was perfect or
  can be blindly applied everywhere
• Imply that we have answers for all the questions

4
Our motivation to write this paper

• Why power rationing? Boring, not forward looking?
• We realized how important it is hands on
  experience
• And that there are sensible, and not so sensible
  ways
• Highlight results, show mechanism, and describe
  the political process non linear, haphazard,
  unpredictable
• Compare with what other countries have been doing
• Realized how important demand side management was
• California interruptible contracts, very
  minor black-outs
• Japan embedded culture or energy savings
• Brazil success in the developing world
• Most importantly help other countries facing
  similar crisis

5
Why is sensible rationing important?

• Power shortages and scarcity are becoming more
  frequent - and systems less resilient
• Oftentimes leading to rolling black-outs the
  worst possible way to allocate a scarce good
• Supply driven, oil based emergency solutions have
  become increasingly expensive
• Interventions on the demand side have not been
  fully explored
• Poor crisis management may jeopardize
  sustainability

6
Capacity scarcity becoming widespread with a
weeping wave in 2001-2002 (just a sample)
7
Apparently, situation has not improved

• 2005
• Nicaraguans have faced power outages of up to
  eight hours (09/20)
• Taipower kicks off 4th edition of power rationing
  program (09/28)
• Bangladesh Emergency meet fails to make
  rationing decisions (10/14)
• DR threatened by energy crisis that could doom
  recent years of economic growth and democratic
  development (12/27)
• 2006
• Severe Power Rationing in Haiti (01/16)
• Tanzania Power Rationing Starts Today (02/02)
• Power shortages and black-outs continue to plague
  Indias major cities (02/03)
• South Africa must race to avert energy crisis
  (02/28)
• Vietnam may face power crisis (03/02)
• Uganda Kampala Power Struggle in a Powerless
  City (03/06)
• French Polynesia Power Rationing Causing
  Problems, Protests (03/0 8)
• Zanzibar starts partial power rationing (03/09)
• Zimbabwe Massive Power Rationing in the Horizon
  (3/20) 

8
Types of shortages and causes

• Energy Constrained Systems lack of MWh
• Poor rainfall Brazil, Norway, New Zealand
  (2001), East Africa (2006)
• Poor operations planning reservoir management
  Tanzania, Brazil
• Conflicting uses for the water Lake Victoria
• However, not necessarily hydro driven
• No money to buy fuel, despite abundant thermal
  capacity (DR)
• Curtailment in fuel supply (e.g. Chile, importing
  gas from Argentina)
• As a result frequent rotating black-outs, any
  time of the day or week
• Capacity Constrained Systems lack of MW
  (peaking capacity)
• Not sufficient generation capacity to meet peak
  load (or fuel) (Kenya)
• Low reserve margins and-or unreliable system,
  high LOLP (SA)
• G cartelization to reduce available capacity
  (California)
• Poor pricing leading to peaks, low load factors
  (MENA countries)
• Energy constrained systems where reservoirs have
  been over depleted (loss of head) both energy
  and capacity constrained (Tanzania)

9
Case studies and best practices
10
Best practice the case of Brazil

• A self-rationing, demand driven scheme
• In a textbook, overly depleted, underinvested
  power system
• Quotas were assigned to individual customers
  (20-25 reduction target)
• Honest perception of crisis and massive
  educational campaign
• Penalties and incentives mostly of pecuniary
  nature and linked to SRMC (about US 300/MWh)
• No single black-out or brown-out -- 20
  reduction, over 8 months, almost country-wide
• At least 1-1.5 of GDP saved (vis-a-vis rolling
  black-outs)
• An efficient safety net fostering engagement from
  the poor
• Not everything rosy but a successful case, with
  many lessons learned

11
Causes and extent of the 2001 crisis

• The most serious energy crisis in recent history
• Affecting 80 of the network, including the
  heavily industrialized and populated Southeast
  region
• Hydro dependent system (gt90) has been strained -
  beyond capacity after several years of below
  average rainfall
• Capacity expansion had not kept up with market
  growth
• Some experience in the past in dealing with
  similar situations but not to this extent and
  magnitude

12
The crisis was not a surprise - 2001 only a
wake up call
13
No problem Rains will come. Perhaps a little of
moral suasion would help
14
This time God was not Brazilian
15
How the government approached the crisis -
competing rationing proposals

• Proposal I Rolling Black-Outs or Apagão
• Geographical regions disconnected on a rotating
  basis, according to a pre-agreed schedule
• Priority loads preserved e.g. hospitals,
  police, etc.
• Apparent advantage immediate implementation and
  100 effectiveness
• Proposal II Quota System
• Allocate quotas to customers, taking into
  account historical consumption and defining
  saving targets baseline consumption
• Enforcement for those who exceed the quota
  financial penalty, threat of disconnection, or
  both
• Apparent disadvantages complex, subject to
  complaints, demand elasticity was unknown
  possibly jeopardizing overall effectiveness

16
Subject was far from consensual

• May 08 ESP - Rationing will have no penalties,
  FHC decides
• May 09 GM Abradee advocates rolling
  black-outs, asap
• May 10 FSP Not even hospitals will get rid
  off apagões
• May 16 JB Abradee insists in rolling
  black-outs to achieve rationalization targets
• May 18 Plan is announced. Quotas with price
  signals. Effective 6/1
• May 19 FSP Energy crisis leaves economists
  lost
• May 21 GM Editorial Plan is based on
  punitive, unfair logic. Government missed
  opportunity to have society as an ally. A
  disastrous plan.
• May 26 FSP Political Leader. The government
  should revisit our proposal to have rotating
  holidays as a civilized way to share costs
  imposed by rationing. There will be 856,000 jobs
  lost (FGV)
• May 28 Globo Utility CEO. There is still a
  strong risk of rolling black-outs. My wife does
  not know how to conserve energy.

17
A quota system was selected Intellectual
leadership or serendipity?

• Quota based on a three-month average of 2000
  consumption (May, June, and July)
• A target reduction differentiated by customer
  group
• Quotas for industry differentiated by segment
• Plus or minus adjustments treated on a
  case-by-case basis (e.g. new customers)
• Financial penalties for those who exceeded the
  quotas
• Bonuses for overachievers
• Possibility to exchange quotas in the secondary
  market
• Signals linked to the energy price at wholesale
  level
• Disconnection threat for those who did not meet
  targets

18
Reduction targets and safety net for the poor
19
Implementation of rolling black-outs would have
been more complex than alleged

• Networks are inter-meshed - essential loads
  40 of feeders
• Intra-day load shift weakens the effectiveness
  of the scheme
• Those two factors would entail a disproportionate
  burden on some customers (10 to 16 hours/day !!)
• Manual nature of disconnection operations,
  difficult to follow a precise timetable
• Lack of energy - melancholic urban chaos
• Traffic lights and crime rates in large cities
• Hurting some manufacturing activities with
  continuous processes

20
Load response was fast and effective without
black-outs or brown-outsMW Peak Consumption
21
Residential energy savings beyond Government
targets
22
The beauty of self-rationing - the customer
decides how to save energy
23
Significant energy savings have been achieved

• Total savings 26 TWh (4 Regions), 13,000 MW
  peak
• For a total national consumption of 284 TWh
• Geographical distribution of savings
• North 0.8 TWh South-East 19 TWh
• North-East 4.7 TWh Center-West 1.5 TWh
• Results of PROCEL National Conservation Program
  (1994-2000) 10.7 TWh, 640 MW peak (2000)
• Savings continued after rationing measures ceased
  
• Overall 2002 consumption similar to 1998 levels
• Average residential 1994 levels

24
Residential customers have indeed changed their
habits, on a permanent basis

• Average individual consumption 1994 levels
• Average consumption pre and post rationing
• SE from 199 kWh/mo to 145 kWh/mo
• NE from 113 kWh/mo to 85 kWh/mo
• 91 of households changed consumption habits
  during rationing from those 65 still maintain
  savings
• Energy efficiency is now part of the decision
  making process to buy appliances
• 8 before crisis
• 58 after crisis

25
Alleged losses to industry players were about R
11 billion more than US 5 Billion

• Losses to Generators
• (In R Billion)
• Revenue Loss R 0.8
• Buy-Back R 3.3
• Non-controllable () R 0.9
• Total R 5.0
• () Not directly related to rationing

• Losses to Distributors
• (In R Billion)
• Cost Reduction R 0.8
• Foregone Annex V
• Revenue Loss R 5.3
• Non-controllable () R 4.6
• Total R 10.7

26
Employment rates remained relatively flat
27
GDP kept growing overcoming the multiple 2001
crises
28
Industrial production grew in the period
29
Can this experience be replicated in other
places?

• Certainly yes but it requires a committed
  government and leadership
• The mechanism in Brazil had some bells and
  whistles but even a very simple scheme may
  produce significant results
• Initial quota assignment
• Deviations priced at SRMC (e.g. cost of emergency
  generation)
• Brazil was an energy constrained case capacity
  constrained have some nuances
• Price signals should be conveyed on a real time
  basis
• But real time metering rarely exists
• Time of use rates and a good dialogue with
  customers may be helpful
• What about commercially distressed power systems?
  
• If price signal does not get to the customers,
  system will not work
• However, other demand side and energy efficiency
  initiative may still be applied
• For example, South Africa is now deploying 7
  million CFLs to deal with crisis
• My colleague Ashok will explore this issue in
  more detail

30
Possible extensions to non-crisis situations

• Post California crisis A lot of research,
  piloting, action - Very promising use of demand
  response to make energy more affordable and
  reliable
• Concept of having a two part system (fixed and
  variable) is appealing, as it creates response to
  prices (Fi-Va) ex-ante negotiated quotas
• Independent system operators establishing
  protocols and agreements with large customers, to
  entertain demand response to price and critical
  conditions
• Combination of price-rationing and quantity
  rationing having a vast impact on peak

31
Lessons learned 1

• Black-outs worst way to ration
• Brazil showed that other ways work
• Price signals help bridge supply-demand gaps
• Rationing program has to be tailored
• Capacity and energy shortages are different
• Plan in advance, well before need to ration
• Have good warning signs to overcome denial

32
Lessons Learned 2

• Explore permanent, quota-like tariff
  arrangements
• Put someone in control to deal with crisis
• Protect the poor with a safety net that
  encourages efficiency
• Do not socialize gains and losses
• Honor contracts