Proposition is made that Green IT is not a technical problem, but an economical problem to be solved.
Presentation as follow up on the publication of the book Greening IT ISBN 8791936020
2. Setting the Stage
As organizations increase their
offerings of digital services, the
demand for communication,
computing and storage capability
increases.
At the same time:
The purchase price of a new (1U)
server has been exceeded by the
capital cost of power and cooling
infrastructure to support that server
and will soon be exceeded by the
lifetime energy costs alone for that
server
Business demands drive the growth of IT energy usage
3. Drive Towards
Energy
Efficiency?
Is there a need to do something about
these rising energy needs of data
centers? What about:
Increasing energy costs for
organizations;
Increasing strain on the
existing power grid;
Increasing capital costs for
expansion and construction of
data centers;
Regulations, standards and
compliance;
Corporate reputation
But are we taking action?
You could say there are technical, financial, legislative, compliancy, and
moral triggers to act and cut down energy usage.
4. Why Wait?
There are many technical solutions to
improve energy efficiency. So what’s
missing here, why is IT not so much
greener than it is today?
Is one of the main reasons that IT is
not green at this moment at least as
much due to perverse incentives?
Shouldn’t we incorporate economic-
behavioral analysis to Green IT to
explain this current failure better?
Business demands drive the growth of IT energy usage
5. IT at Work,
The Risk of
Incoherency
The components of power, cooling, air
handling last a longer time than IT
infrastructure.
As an assurance, frequently an excess
of devices are purchased to
guarantee capacity in extreme IT
load scenarios.
Responsibility of using a data center
falls across projects, IT departments,
business departments, corporate
real-estate and (depending on the
sourcing model being used) other IT
organizations.
With conflicting interest and siloed decision making
energy inefficiency becomes the rule
6. Energy at Work,
The Risk of
Flexibility
In the energy supply chain there are
several actors. Organizations that
generate power, transmission network
organizations, organizations for
distributing power and companies that
sell electricity to customers.
A very specific element of this
infrastructure is that there is no storage.
Demand and supply must be in
equilibrium, else there is the risk that
this infrastructure shuts down.
Because of these specific properties
capacity management must be
coordinated between all involved
organizations.
The electrical power infrastructure is inelastic.
7. Supply Chain
The IT infrastructure is basically a value stack. A supply chain of stack
elements who acts as a service component (People, Process and IT that adds up to
an IT service). In this supply chain there are several actors.
In the data center two critical infrastructures, the IT infrastructure and the
electricity infrastructure are tied together. The current power grid infrastructure is
not flexible and has low energy efficiency.
The current “IT grid” infrastructure has low energy efficiency and poor energy
elasticity
8. DC Supply
Chain
A data center has, from a power
perspective, a supply chain that
consists of four large building
blocks: the IT infrastructure, the
primary power supply (UPS, PDU,
etc.), the secondary support supply
(cooling, generator, air handling)
and the tertiary support supply
(lighting, office space).
To improve the energy efficiency of
existing data centers, as well as
making decisions on new data
centers there are some metrics
being used.
These metrics and processes will
help determine whether the
existing data center can be
optimized
IT Load
Servers
Storage
NetworkCooling
Generator
Switchgear
UPS
PDU
Cabling
etc.
Data center
Physical infrastructure
Power to
data center
Power
to IT
Power path to
IT
Power path
to secondary
support
In parallel
with IT load
In series
with IT load
IT infrastructure
Total Facility Power
IT Infrastructure Power
PUE = IT Infrastructure Power
Total Facility Power
x 100%DCiE =
Useful Work
Total Facility Power
DCP =
Power path
to tertiary
support
Lighting
etc.
9. Supply & Demand
Power
Generation:
10 years to fix
Transmission:
10 years to fix
DC Building:
1-5 years to fix
IT infrastructure:
1-3 years to fix
CO2 emission
Capacity
constraints
Conversion
Efficiency
Capacity
constraints
Stranded capacity
Capacity
constraints
Stranded capacity
Efficiency
10. Which Energy Efficiency
Are we talking about?
Energy savings; emphasize on (absolute) reducing energy consumption because of
the need to reduce consumption of primary energy resources (oil and gas) caused
by the Oil Crisis.
Energy conservation; emphasize on (absolute) reducing energy consumption
because of the need to reduce consumption of primary energy resources (oil and
gas) because they were regarded as in danger of exhaustion.
Energy efficiency; emphasize on becoming increasingly efficient in the usage of
energy whilst economic growth can cause continuing use of more energy.
Energy productivity; emphasize on sustainable development because of the
scarcity of energy resources and the interest and concern related to climate change
and carbon dioxide emissions.
11. The Effect
The effect that increases in energy
efficiency raise energy consumption
is known as the Khazzoom-Brookes
Postulate.
As stated by the economist Saunders
(1992), this is explained by the micro
level in- creases in energy efficiency
leading to lower costs of energy, and
on the macro level side increases in
energy efficiency leads to increased
economic growth.
The Khazzoom-Brookes Postulate is a
special case of what in economics is
called the Jevons paradox, increases
in the efficiency of using a resource
tends to increase the usage of that
resource.
12. The Gap
It looks like that Green IT is more
about sustainability and thus energy
productivity then energy efficiency.
Considering the Climate change
issue, the goal of economic
performance instead of economic
sustainability looks rather short
sighted.
To answer this question we can look
at a micro, meso and macro level of
this issue or to put it an other way
on the level of a firm or
organization, the market and
society.
The gap that exists between actual and optimal energy use rise the question
how to define the optimal level of energy efficiency.
13. Optimize
Micro level - Firm/organization optimum; indicates the amount of energy
efficiency that might be expected to occur under current market conditions and
market behavior
Meso level - Market optimum; indicates the amount of energy efficiency that can
be achieved if all technologies that are cost-effective from a consuming
organization point of view were implemented.
Macro level - Social economic optimum; describes the amount of energy efficiency
that would be achieved if all technologies that are cost effective based on a social,
rather than a private perspective (by taking externalities in to account) were
implemented.
Macro level - Hypothetical optimum; represents the maximum amount of energy
efficiency that would be achievable through technology diffusion if all
technically feasible technologies were used without regard to their cost
acceptability of certain stakeholders.
To reach one of these energy consumption optimization levels, barriers must be
dismantled and be removed. Which kind of barriers must be eliminated depends
on the optimum one tries to achieve.
14. Rationality?
A central idea in economics is that in the free
market, actors behave selfishly and try in
rational ways to reach their goals. This
market leads as if by an invisible hand to
economic efficiency.
The behavior of the actors, and the efficiency
and performance of the accompanying
markets, should lead to energy efficiency as
a function of prevailing prices.
Rational decision making requires that the
decision maker has all the information he
needs and that he is knowledgeable about
the consequences of possible decisions.
In real life not everything is known, the
decision maker is limited by the amount of
time and resources he has to make decisions,
so rationality is bounded.
15. Information
Asymmetry
If all the parties that are involved in a
transaction don’t have access to the
same relevant information, then we
are talking about information
asymmetry. This creates an
imbalance of power in a transaction.
Where one party is trying to motivate
another party to act on ones behalf,
this is also known as the principal-
agent problem and can lead to adverse
selection and moral hazard.
Bounded rationality, information
asymmetry and opportunism leads to
strategic behavior
16. Strategic
Behavior
Strategic behavior with rules/policies;
parties interpret policies, legislation or
contracts in their one gain and act
accordingly boundaries.
Strategic behavior with information; parties
are selective with sharing information with
other parties because they think there is a
gain in doing so.
Strategic behavior with prices and
quantities; parties setting their services and
product prices accordingly to the market
power of the other parties.
Strategic behavior with bottleneck facilities;
parties abuse their ownership or control of
an essential component of a system,
through which all service products must
pass to reach the ultimate buyers.
17. Market
Failure
In some way there is a lack of
sufficient incentives to create an
effective and potential market in
energy for IT, and the nearly
nonexistence of this market results in
the loss of efficiency in energy usage.
This market failure has three aspects:
Externality
Tragedy of the commons
Property rights
18. Costs
Energy efficiency improvements are
burdened with discussions about
costs.
In particular, it is a discussion
whether all the costs are taken into
account for a proper costs evaluation.
There is also a discussion about
where to draw the line between
transaction costs, switching costs,
hidden costs or even production
costs.
This forms one of the barriers in the
diffusion of energy efficient
technologies.
A proper overview of all the costs that are involved in diffusion of
technologies could help to explain this barrier.
19. Risk
Appetite &
Innovation
Risk Appetite is the amount of risk
exposure that the individual, group or
organization is willing to accept. This is
influenced by:
The used business models and the
derived incentives,
The (perceived) flexibility of the IT
service stack to make switches
(interdependencies),
The estimation of the (perceived) costs
that are involved in making a transition,
and
The risk of outages when making
transitions.
If there are uncertainties about risks when using new solutions they will
form a huge barrier in the diffusions of new energy efficient technologies.
20. Yes, we can
change
In cutting down energy usage and energy costs, the focus is mostly on technical
measurements, but figures show that this is not enough.
There is a disconnection between the environmental sustainability and the
economical sustainability.
Applying economic-behavioral analysis to IT energy usage and dependability
shows that it often explains energy efficiency failure better.
Systems are often wasting energy because the people who manage them, or who
could fix this, have insufficient incentives.
Energy usage is for most involved parties an externality.
21. Optimization
level
Issues
Decision
makers
Stakeholders Goals Criteria Scenarios Actions
Firm
Organization
Market
Social
economic
Solution
framework
How much should be spend of efficient energy usage?
That depends on which optimum you have in mind. These levels of optimal energy
usage can form a framework for solutions.
Because asymmetric information leads to many behavioral aspects, the framework
must explicitly take decision makers and stakeholders into account.
22. Organization
optimum
The issues to be solved over here are
more or less technical and procedural
coordination issues and motivational
issues based on lack of information:
Interdependency of the service stack
Cost management
Risk management
The goal to achieve is:
Eliminate barriers considered by
economists as rational, achieve
operational excellence. That is the
amount of energy efficiency that might
be expected to occur under current
market conditions and market behavior.
23. Market
optimum
The issues to be solved over here, ask
for active interventions and the success
rate depends a lot on the actual power
the organisation has in the market:
Strategic behavior and information
asymmetry
Split incentives
The goal to achieve is:
Eliminate market failures, which can
pass a cost-benefit test. That is the
amount of energy efficiency that can be
achieved if all technologies that are
cost-effective from a consuming
organization point of view, were
implemented.
24. Social economic
optimum
Issues to be solved over here, because
of the (environmental) externality, ask
for active interventions of the
government in terms of policies and
regulations:
• Market failure
• Strategic behavior and information
asymmetry
The goal to achieve is:
Eliminate market failures, including
environmental externalities, which can
pass a cost-benefit test. That is the
amount of energy efficiency that would
be achieved if all technologies were
implemented that are cost effective
based on a social, rather than a
consuming organization perspective
(by taking externalities into account).
25. Read all about
it
The book, Greening IT (ISBN
8791936020) aims at promoting
awareness of the potential of Greening
IT.
The foreword of the book is made by,
European Commissioner for Climate
Action, Connie Hedegaard.
Greening IT is a international, non-
profit, collaborative effort of a group of
writers to cover as many areas of Green
IT.
The chapter “Why Green IT is Hard - An
Economic Perspective” is my
contribution to this book.
See http://greening.it
Hinweis der Redaktion
According to the common view, Green IT comes down to implementing technical measures. The idea is that, given better power management of equipment in the workspace (such as laptops and PC ’ s), more efficient power usage of servers, storage and network components, virtualization of servers, better power and cooling management in data centers, the problems can be solved. But is this really true? The reason IT is not green at this moment is at least as much due to perverse incentives. Green IT is about power and money, about raising barriers to trade, segmenting markets and differentiating products. Many of the problems can be explained more clearly and convincingly using the language of economics: asymmetric information, moral hazard, switching and transaction costs and innovation. Green IT is not a technical problem, but an economical problem to be solved.
Additional demand for computing capability can only be met by increasing the processing capacity of servers within data centers. Along with greater computing capability, businesses have increased demand for storing digital data, both in terms of amount and duration due to new and existing applications and to regulations.
“ It looks as if various factors prevent energy consumers from taking actions that would be in their private self-interest to do so, that is, would result in the provision of services at lower cost because of diminished energy costs. ”
“ for a lot of data centers, the criticality of their function to the business just prohibits downtime and inhibits facility managers from making major overhauls to realize energy-efficiency improvements”
Zooming in and looking closer to this value stack or supply chain we see not only the expected functional interdependency, but also several hard, unwanted, dependencies between the stack elements from a technical, procedural and organizational perspective. Because of this interdependency, inefficiency in the IT infrastructure has consequences for the energy infrastructure and vice versa.
Based on ‘ The energy gap ’ Jaffe and Stavins (1994), here we define four potential levels of optimising energy usage.
Adverse selection is taking place when ‘ wrong ’ choices are being made, because of asymmetric information, inefficient transfer of information between producers (agents) and consumer (principal). For example reliance on old information may pro- long the use of outdated or suboptimal technologies, even though this is economically inefficient. This can limit the usage of energy efficient products. Moral hazard is the fact that the principal lacks information about the performance of the agreed-upon transaction or lacks the ability to retaliate for a breach of the agreement. Therefore the agent may behave differently from the way it would behave if all the information about the transaction performance was available.
Externality : a party makes a choice or transaction that has an effect on other parties that are not accounted for in the market price. In such a case, prices do not reflect the full costs of production or consumption of a product or service. Tragedy of the commons : This concept refers to the situation that if people ignore the scarcity value of the commons, they can end up over harvesting a resource and leading to environmental degradation. Where in this case energy is seen as the common property, which usage is non exclusive. Property rights : If a party doesn ’ t have the right controls over the use of their commodities, exclusive use and/or the delegation of use and the related costs of doing so, this will result in inefficiency.