Electricity load-shedding in Pakistan: unintended consequences,

opportunities and policy recommendations

Hussain Kazmi1, 2, Fahad Mehmood3, Zhenmin Tao1, Zainab Riaz3, Johan Driesen1
1
ESAT, KU Leuven, Belgium
2
Enervalis, Belgium
3
Lahore University of Management Sciences, Pakistan

Abstract
Over the last decade, supply-side constraints have resulted in widespread electricity shortage in
Pakistan. At its peak, this amounted to over a 7 GW supply-demand gap and caused the electricity
grid to be offline for vast swathes of population for many hours daily. Despite major supply-side
investments acute shortages persist and a large percentage of relatively affluent households, estimated
in millions, have countered this by investing in self-generation and battery storage technologies
(usually lead-acid batteries because of their low cost). This paper summarizes the impact of this
backup technology on the broader energy system in terms of efficiency losses for households and
contribution to low-voltage grid congestion. Research findings suggest that the low efficiency of these
backup systems has caused annual losses of around 3-4 TWh for the electric grid in Pakistan as well
as overloading of transformers and frequent supply-demand imbalances. However, the widespread
adoption of these backup systems has also created an entire ecosystem which can enable massive
demand side management and provide the framework for a future smart grid in Pakistan. Besides
evaluating the opportunities, possible policy measures the government should undertake to enable this
transition are also discussed

Introduction
Electricity supply and demand must be balanced at all times for stable grid operation (Kundur,
1994). This balancing has to take place instantaneously for the grid to function properly. While a
robust grid is usually capable of dealing with minor discrepancies, major imbalances between supply
and demand lead to undesired frequency shifts and, in extreme cases, cause equipment failure and
large-scale blackouts (Amin, 2005; Andersson, 2005; Baosheng, 2003).
In developed economies with stable electric grids, electricity blackouts are avoided by
anticipative contingency planning for security of supply. This planning takes place over multiple
spatial (transmission and distribution grid levels) and temporal scales (long- and short-term time
horizons) (Boston, 2013). While this does not necessarily preclude a blackout in exceptional
circumstances (for instance in case of natural disasters leading to infrastructure damage etc.), proper
planning ensures that there are no recurring electricity outages (Helm, 2002; Bayod-Rújula, 2009).
However, the situation is reversed in many developing countries which have weak electricity grids
and electricity demand often exceeds supply. Coupled with increasing levels of urbanization and
electrification and very limited data gathering capabilities to support long term decision making,
balancing supply and demand is often an extremely complicated issue for the grid operator.
Over a billion people today still do not have access to electricity (IEA, 2017). In India alone,
there are 239 million people without access to electricity. Neighbouring Pakistan and Bangladesh too
suffer from low electrification rates with about 25% population without electricity access, as do many
countries in Sub-Saharan Africa and, to a lesser degree, Latin America (IEA, 2017). However, while
efforts to electrify these people continue, there is a large percentage of people in these countries who
have access to electricity, yet they suffer from periodic blackouts. The focus of this work is
addressing electricity reliability issues for these households, estimated to be almost another billion
worldwide (IEA, 2017).
 The historical solution to address security of supply has been to enhance supply side investment
and grid reinforcement. These are however expensive propositions and can take many years to
implement. In the absence of established demand side management (DSM) programs, the most
common recourse for system operators in this regime is to either reduce the operating voltage (which
in turn reduces the demand) or, in more extreme cases, to operate the grid with recurring blackouts
(Warnock, 1986). These blackouts are seen as a means of balancing supply and demand whereby
rotating supply to different regions can effectively mean most customers receive electricity at least
some of the time.
It is well documented that recurring electricity outages cause enormous economic losses (Pasha,
1989; Adenikinju, 2005; Kessides, 2013). However, they also spawn a multitude of unintended
consequences. Foremost amongst these is the widespread adoption of backup generation systems by
different households (Ahmad, 2016). These backup systems are frequently in the form of battery-
inverter systems or diesel generators, where the battery-inverter system is by far the more common
because it is less intrusive and is cheaper to both acquire and operate. However, the true operational
costs of such a backup system are hidden because building occupants are, in general, oblivious to their
actual energy needs and the real efficiency of the backup system. This imperfect information system
can lead people to make uninformed decisions which do not make economic sense.
The battery-inverter systems operate at an efficiency which is well below 100%h leading to a
systemic loss of efficiency which adversely effects not just the households but also the energy utilities
and the economy in general. It is well documented that electric grids in developing economies are
plagued by technical and non-technical losses (Antmann, 2009). In Pakistan, for instance,
transmission and distribution losses already amount to almost 18% of the annual generation (NEPRA,
2016). Widespread use of such battery-inverter systems further exacerbates the situation. Mass
adoption of these backup systems can also contribute negatively to local low-voltage grid stability
because of synchronized charging cycles. All these factors combine to make an already bad situation
worse. At the same time, the policy decision to implement rolling blackouts translates to punishing
the economically vulnerable while driving up prices and demand of electricity for the more affluent.
The grid too suffers from reduced longevity and efficiency. it is anticipated that intelligent policy
changes can help find a better solution to the problem of supply-demand imbalances in developing
economies experiencing rapid urbanization and electrification.

BackgroundIn this paper, the aforementioned problems were analysed in the context of the
electric grid in Pakistan. The case of Pakistan is unique in the sense that a majority of its population
without access to electricity is above the poverty line. So, for instance, while the overall electrification
rate in Pakistan is similar to Bangladesh, the proportion of people without access to electricity above
poverty line is inverted (i.e. 75% in Pakistan vs. 25% in Bangladesh) (IEA, 2017). This means that the
majority of people without electricity access in Pakistan can afford it but simply do not have access to
it due to structural issues. While this supply-demand gap reached as high as 7 GW in 2012, it has
generally swung between 4 and 6 GW since then (IRENA, 2018).
End electricity sales in Pakistan have risen from only 17.6 TWh in 1985 to 100 TWh in 2017
(NEPRA, 2017), (NEPRA, 2006). This is almost a six-fold increase in sales. In general, supply has
lagged behind demand and two significant factors leading to persistent shortage are the increasing
electrification rates and rising electricity demand in grid-connected users caused by population and
economic growth. Part of this can be explained by increasing population (a higher than two-fold
increase from 92 million in 1985 to almost 200 million in 2017), while part can be explained by
increasing per-capita incomes (an almost six-fold increase from 1000 USD/capita in 1985 to almost
6000 USD/capita in 2017) (World Bank, 2018). More recent years have seen an acceleration in
demand with growth for the last five years turning out to be almost 20%. While existing consumers
have seen their demand increase by 7% during this time, the remainder has largely been driven by
new connections. In the same timeframe, the peak power demand has risen from 25.5 GW in 2012 to
almost 27.5 GW in 2017 (NEPRA, 2017).
The residential sector in Pakistan was responsible for 24% of its total energy demand of 41.98
Mtoe in 2015. Further zooming into this consumption, domestic demand makes up more than 48% of
all electricity consumption (IRENA, 2018), having edged up from 45% a decade earlier (NEPRA,
2006). This electricity demand is highly seasonal as in most of the population centres of the country,
summers are very hot while winters are mostly mild. Furthermore, heating in winters is often carried
out using natural gas. This leads to a large spike in electricity demand during summers to provide
ventilation and air conditioning in residential and tertiary buildings (NEPRA, 2017). As affluence
levels rise leading to greater demand for air conditioning, this problem will worsen in the future.
The purpose and contribution of this paper are threefold. Firstly, the unintended consequences
arising from wide-spread and unregulated adoption of uninterruptible power supply systems to
mitigate electricity outages are highlighted. These systems have a debilitating effect on the energy and
peak power experienced by the electric grid. Secondly, practical insights on how the detrimental
effects of these residential batteries can be reduced is presented. Lastly, the appropriate policy choices
to provide the roadmap for a future electricity grid are discussed.

Unintended consequences
Widespread electricity load shedding has led to mass adoption of backup systems which usually
employ lead acid batteries (Arshad, 2017) (Ahmad, 2016). In common parlance, these are often
referred to simply as Uninterruptible Power Supply (UPS). These backup systems are usually
considered as short-term storage solutions to bridge a single electricity outage event. While other
alternatives such as diesel generators are available in the market, UPS systems have become pervasive
because of their perceived low cost, and seamless and noiseless operation. However, this adoption has
led to some unintended consequences, some of which are highlighted in this section.

Loss of efficiency
The charging efficiency of lead acid batteries is typically assumed to be around 85% (Stevens,
1996). The overall efficiency, obtained by combining charging and discharging efficiency, is usually
lower. In our experiments, conducted in real-world conditions, this was however not the case. More
concretely, two test batteries were installed in residential settings where they were allowed to charge
and discharge according to the availability of the electric grid and household loads. According to
readings from a smart meter, both were found to have a charging efficiency of less than 60%.
Combined with the discharging efficiency, the overall efficiency dropped below 50% for both cases.
This result, while disappointing, correlated well with existing studies which show that overall
efficiency of commonly used UPS systems can be as low as 50% (Ahmad, 2016). Previous studies
were however conducted with resistive loads; in real life applications, the efficiency turned out to be
even lower because of presence of inductive loads and temporal degradation of the battery.
Besides the poor quality components of the battery, another reason for this low efficiency is its
control strategy. The normal operation of a UPS leads it to charge the battery as soon as electricity is
available from the grid and discharge as soon as the grid goes offline. This operation mode disregards
two further important considerations in the charging and discharging of lead acid batteries. The first is
a nonlinear charging profile which leads to a decreasing efficiency with increasing state of charge of
the battery (Stevens, 1996). The second is that repeated cycling to extremely low states of charge is
detrimental to the battery’s longevity, especially for lead acid batteries (Young, 2013).
The quantification of efficiency loss for households is a complicated endeavour because it
depends on both the load profiles and the outage patterns. Nevertheless, our simulations conducted for
a wide variety of outage conditions and demand profiles show that additional losses are on the order
of 10 to 20% of annual consumption for urban, relatively affluent households consuming between 5
and 10 MWh/a. Concretely, this comes out to be between 500 kWh/a and 2000 kWh/a. These results
mirror those in other recent studies, which have placed this number between 600 kWh/a and 1800
KWh/a (Ahmad, 2016), (Arshad, 2017). At a national level, the annual losses caused by use of these
batteries can be extrapolated to be 1-4 TWh, i.e,1-5% of the total end electricity demand (Ahmad,
2016) (NEPRA, 2017).

Grid issues
The control strategy in its present form, besides reducing operational efficiency and longevity,
also leads to congestion in the local low voltage grid. This is a necessary corollary to an already
stressed grid and mass proliferation of new synchronized electrical loads. In many developed
countries, where heating networks and transportation are currently being electrified, similar
challenges are arising (Clement-Nyns, 2010). The situation is worse in developing countries where
existing infrastructure is usually unable to cope with even existing demand. Indeed, according to the
grid regulatory authority in Pakistan, both the high and low voltage grid stations are already heavily
overloaded. Recently, at the low voltage level, over 31% of all 11 kV feeders were found to be
overloaded as were 15% of all distribution transformers (NEPRA, 2017). These numbers vary
significantly by region in the country from 10% to 25% depending on the time of the year. Summers,
when both grid load and outages are at their highest level, also correspond to the time period when
transformers are most overloaded, thereby worsening the situation for end consumers.
To test the hypothesis that additional loads in the form of electric batteries cause transformer
overload, this research simulated a low voltage electricity grid with a transformer rated at 25 kW
using PYPOWER, a software tool for designing and optimizing electric grids. This transformer was
designed to serve an urban neighbourhood of 15 houses. Without any batteries in the neighbourhood,
this transformer was only occasionally overloaded. Furthermore, this overload did not exceed the
nominal capacity substantially. However, with only one third of the households equipped with 1
kWh/1 kW battery-inverter systems, the transformer was almost continuously overloaded during the
summer months. For significant amounts of time, the transformer was overloaded to over 170% of its
capacity which can severely reduce its longevity. In most affluent neighbourhoods, household UPS’s
are dimensioned to be larger than 1 kWh / 1 kW as shown in Fig. 1 obtained through a survey in
Lahore University of Management Sciences (LUMS). Furthermore, the rate of proliferation is also
much higher than 33% (Pasha, 2013).
In addition to causing grid congestions, these batteries have also been demonstrated to produce
harmful harmonic distortions in the power system. This harmonic distortion can lead to overheating of
transformers, stressing of power factor correction capacitor banks and reducing the lifetime of
electrical appliances (Ahmad, 2016).

Dimensioning the backup system

In a survey conducted at Lahore University of Management Sciences (LUMS), where over 50
people participated, the annual household energy demand turned out to be only weakly correlated (<
0.5) with the installed backup system dimensions. It is straightforward to infer from this that backup
system dimensions are guided not by actual demand, but fear (of lost load) and affluence levels. The
prevalence of a 2 kW inverter system (regardless of household demand) also signifies the systemic
lack of information that most households have to contend with.
In the long run, such miscalculations can end up being costly for the household because of their
higher upfront and replacement costs, and the energy losses they entail. 2 kW inverters are, in
practice, usually paired with 2 kWh batteries. Unless there is a need for this larger battery driven by
actual demand, these higher capacity batteries increase losses and are much more expensive in terms
of upfront costs. As the life expectancy of a lead-acid battery with frequent deep discharges is no
more than 200 – 300 cycles, this requires repeated replacement (often multiple times during the same
year) (IRENA, 2017). The batteries are therefore problematic not just for the loss of efficiency but
also the investment costs.

Policy recommendations
While the policy of endemic load shedding has crippled industry and adversely affected
households, it has also brought about some unprecedented opportunities. These can provide a
roadmap away from the supply-side investments the government has focused on, and towards a
greener, smarter energy grid in Pakistan. In a market marred by lack of transparency and high barriers
to entry, regulatory and policy intervention can help materialize these opportunities. In the following,
we discuss some of these possibilities in detail.

Ensure information dissemination and implement quality control

As discussed previously, ample evidence exists that the efficiency of widely available UPS
systems is far below international standards. Furthermore, because of the information asymmetry,
most consumers have no knowledge of this low efficiency, or how it affects their energy consumption
(or electricity bill). Finally, the systems are dimensioned in a random manner which further increases
losses and cost to the users.
An obvious policy recommendation is to implement strict quality control checks on the products
being sold and marketed to the general public. Since low efficiency affects not just the households but
also the grid at large and the energy imbalances in the country, it is imperative to ensure that too much
electricity is not being wasted as avoidable technical losses. These quality controls are only the
starting point however and should pave the way for energy labelling of all backup systems. These
labels should be designed to clearly identify the economic impact of the lifetime energy consumption
of such devices. This is important because right now cheaper but less efficient systems are crowding
out more efficient ones due to endemic information asymmetry. The labelling process should follow
lessons learnt from energy efficiency labelling standards already setup in different domains in
developed countries (e.g. for energy consumption in buildings) (Brounen, 2011) (Gram-Hanssen,
2007).
It is also highly recommended to minimize the current information asymmetry by introducing
tools to at least inform the users about the consequences of using such inefficient batteries. Energy
labelling is a necessary but insufficient part of this information dissemination process. Electricity
utilities need to be more proactive in communicating this information to the end users. While it is
unlikely that households will engage in active control of the charging cycles of batteries, they might
be able to make more informed decisions about the dimensioning of the backup system. Such
information dissemination can also go beyond the dimensioning of just electrical batteries and focus
on reducing overall household demand.

Invest in smarter infrastructure

Smart meters, or advanced metering infrastructure, are a prerequisite for modern smart grids.
Collecting data can not only identify and help with forecasting demand patterns, it can also help cut
down on technical and non-technical losses by identifying hotspots of such activity. This information
can be used both for operational optimization and infrastructure upgrades. A number of smart meter
initiatives are already in place in Pakistan, however they need to become more transparent and take
central stage to modernizing the energy infrastructure. Another key component of the smart meter
infrastructure is the widespread adoption of modern telecommunication infrastructure which is slowly
being put into place. Such fine-grained data is also essential to enable more advanced smart grid
services.
 Smart grids comprise of many elements, however a commonly agreed upon use case is DSM.
DSM involves both demand reduction and demand response. Reducing demand can be achieved by
adopting more energy efficient devices (passive measures) or better controllers (active measures).
Responding to price signals or automated direct load control can result in both demand reduction and
response. However, this response would require flexible loads in addition to the classical alternative
of modulating supply to meet fluctuating demand. While the recent surge in interest in DSM has been
led by higher proliferation of variable renewable energy sources in developed countries, the backup
systems in households can provide a valuable source of flexibility to the grid.
Arguably the simplest task for a smart controller is to improve energy efficiency of the backup
system operation by better matching electricity outage patterns with local demand and the battery’s
charging profile. Since the battery has a nonlinear charging profile, operating the battery in a lower
state of charge can help improve operational efficiency. This must be balanced with deep discharging
of battery which can lead to reduced battery longevity or, worse, lost load for the household. Through
simulations, it was found that such a controller can improve the operational efficiency of the battery-
inverter system by up to 20%, which can reduce the household electricity consumption by up to 7%.
These controllers follow the same principles as optimal control for thermal storage (Kazmi, 2018).
Unlike improving energy efficiency, which is a local control problem, improving grid congestion
using electric batteries requires cooperative actions by batteries operating in different households in
the entire neighbourhood. At a local level, the primary goal of this is to reduce transformer overload.
This can be done without inconveniencing households through (automatic) demand response.
Through simulations, it was found that it is indeed possible to reduce transformer overload
substantially through setting up an appropriate reward stream. To test this, the same setup as before
was used where a population of 5 houses were equipped with backup systems of 1 kW / 1 kWh in an
urban neighbourhood of 15 households. The transformer overloading above 170% was completely
mitigated while transformer overloading above 100% was also reduced by almost a quarter. While the
exact amount of overload reduction depends on the transformer’s rating, the availability of batteries
and household demand, these results identify the potential of using smart control to increase the utility
of multiple stakeholders. Reducing harmonic distortion, on the other hand, needs to be addressed
primarily at the manufacturer’s end through quality control checks, as mentioned earlier.
Our estimates place the energy capacity / activation potential of these backup systems at over 2
GWh / 2 GW. Other recent estimates likewise have placed them even higher at close to 3 GWh / 3
GW (Ahmad, 2016) (Pasha, 2013). For an electricity grid with peak production of between 25 and 30
GW, these backup systems represent an enormous amount of untapped flexibility. They also represent
enough capacity and activation potential to cover most frequency response, containment and
restoration requirements in a well-planned and operational electricity grid. For instance, (IRENA,
2017) defines the typical requirement for frequency response and containment as 1 to 50 MW
installations with 1-1.5 energy-to-power (E/P) ratio. This can be achieved through pooling smaller
units, which is definitely the case with household electric batteries. For comparison, the Belgian
electricity system has a fairly similar annual energy demand of 82 TWh and 21 GW installed capacity
(with a peak capacity of 12.7 GW) (IEA, 2016). Elia, the Belgian transmission system operator,
contracts frequency reserve for grid stability at multiple levels. The fastest responding of these, the
primary reserve which has to come online in a matter of seconds, is required to be only 100 MW
(Elia, 2013). This is a fraction of the already installed activation potential in residential batteries in
Pakistan. Many pilots have already demonstrated the value of such automated demand response
(Oudalov A. C., 2007). There is even historical precedent for use of lead-acid batteries as grid
support, albeit at utility-scale. In 1986, a 14.4 MWh / 17 MW battery plant was used to provide
frequency control in the electricity grid of West Berlin (IRENA, 2017).
 Improving energy efficiency directly benefits the household, however reducing grid congestion
or providing frequency response involves multiple stakeholders. Furthermore, reducing synchronous
charging of batteries can only be done through spreading the battery’s charging in time. This
necessarily increases the potential for lost load. As most of the value of reducing congestion or
imbalances is directly accruing to the distribution or transmission grid operator, an incentive
mechanism needs to be setup for the households. This is again only possible in the presence of a smart
metering infrastructure to enable real time monitoring and data gathering.

Update grid codes

Grid codes define the rules and requirements of engagement between the power system and
energy market operation (Christiansen, 2006). They are primarily meant to enable operational stability
as well as security and quality of supply. In Pakistan, there is an urgent requirement for the system
operator to update the grid codes as these were setup in 2005, i.e. before electricity blackouts became
a recurring phenomenon in most cities (NEPRA, 2005). Grid codes are synergistically related to
energy policy: the appropriate level of stringency in grid codes ensures system stability while
ensuring energy policy objectives (e.g. increased electrification, more cost-effective or renewable
electricity generation) are met. Currently, the grid codes in Pakistan provide wide reaching powers to
the system operator, with only a passing reference to ensuring fairness. For better transparency, it is of
paramount importance that the system operator starts providing real (or near real-) time production
and consumption data. This can help with innovations in the energy services being offered and pave
the way towards full-fledged DSM based on market signals (an example for this can be the setting up
of day-ahead and imbalance markets in the mould of many countries with liberalized electricity
markets). It is imperative to revise grid codes in light of lessons learnt in other countries and in
consultation with local stakeholders. As a starting point, these grid codes are required to formalize the
role of low efficiency electric batteries in the entire system.
It is not possible for the system operator to contract this flexibility directly from households at
the moment. No market based mechanism exists in Pakistan which could enable this. A critical policy
recommendation therefore is for the system operator and legislators to setup aggregator and balance
responsible party (BRP) roles in the broader energy framework. While it is infeasible to expect the
entire flexibility activation and capacity of residential batteries to be brought online under such a
scheme, even a fraction would help provide valuable ancillary services to the electric grid. Given the
fast nature of battery operation, this can be leveraged to provide both primary and secondary reserves
as well as local congestion control. However, the nature of this operation means that it can only take
place if (1) it is automated and (2) the corresponding incentive structure is in place. Very recently,
through the Amendment Act 2018, the government of Pakistan has introduced the concept of trader
and supplier of electricity. However, for now these are vague terms and the regulator has rightly noted
that the software and hardware framework as well as the human resource expertise required to
accomplish electricity trading remain missing (NEPRA, 2017).

Devise and implement building codes

Building codes in many countries include specifications on both safety and overall energy
demand. For instance, the European Union has issued a series of directives aimed at reducing the
energy consumption of buildings. These include the 2010 Energy Performance of Buildings Directive
and the 2012 Energy Efficiency Directive stipulating that by the end of 2020, Member States must
ensure that all newly constructed buildings consume ‘nearly zero’ energy (Schimschar, 2011). The
definition of nearly zero was left unspecified which allowed for a degree of freedom, however the
effect has been quite clear. The number of high efficiency buildings in Europe has steadily risen and
is set to steadily increase in the future (Visscher, 2016).
In Pakistan, ventilation and air conditioning makes up for a large part of the domestic load, and
in many cases electricity consumption for a household during the summer months can be 5 times
higher than during the winter months. This additional demand is a major contributing factor in the
peak load-shedding period during summer. A major driver for this increased energy consumption is
inefficient insulation and building design. In a market with severe lack of information, regulation on
building codes can substantially improve energy efficiency. Furthermore, energy labelling and
imposing minimum efficiency requirements for air conditioning and ventilation systems can further
reduce the peak summer demand. In Pakistan, the National Energy Conservation Centre
(ENERCON), in collaboration with Pakistan Engineering Council (PEC), issued a Building Energy
Code of Pakistan Energy Provisions 2011. This is a statutory regulation to ensure building standards
for energy conservation across the country (PEC, 2014). It provides minimum requirements for
energy-efficient design/construction of buildings and applies to buildings and clusters of building that
have: a total connected load of 100 kW or greater, or a contract demand of 125 kVA or greater, or a
conditioned area of 900 m2 or greater, or unconditioned buildings of covered area of 1,200 m 2 or
more (IEA, 2011). The code prescribes better insulations for effective cooling and heating besides
other energy conservation methods. For example it includes prescriptions on: the right amount of
lighting in rooms; applying heat-resistant materials; natural ventilations; types of transformers etc.
The code, if implemented, is expected to result in 25 to 30% of energy savings in the new buildings
and overall savings of 2,500 – 3,000 MW electricity and more than $2 billion a year in foreign
exchange (Sohail, 2011).
However, implementation and awareness of the regulation and its awareness remains a challenge.
No such study has been found that determines or quantifies the effectiveness of the regulation once it
has been implemented. Furthermore, the regulation focuses on tertiary buildings; in a country where
residential electricity demand makes up almost half of total electricity demand, this is a grave
oversight and needs to be rectified with proper legislation. There should also be stringent criteria in
place to monitor energy consumption in domestic as well as commercial buildings to implement the
guidelines for energy efficient buildings. Earlier policy recommendations on promoting smarter
infrastructure in conjunction with building codes will ensure improvements in energy efficiency.
In addition to building level improvements, renewable energy sources in the form of distributed
generation can also play a large part in reducing both the overall supply-demand gap and the peak
demand. As a country with massive solar potential, local renewable generation in Pakistan should be
combined with these battery storage systems to help reduce the load on the electricity grid. Since load
coincides almost perfectly with increased generation (during the summer months), new synergies can
be developed through smarter control algorithms which help reduce both technical losses and
counteract grid interaction. Another possibility is to mandate minimum levels of self-consumption of
renewable energy through the already installed battery-inverter systems for affluent households.

Conclusions
This paper has focused on the unintended consequences of large scale electricity load shedding in
Pakistan. These have included mass adoption of battery-based backup systems which people use to
ensure uninterrupted power supply. The use of these electric batteries has led to additional technical
losses in the form of reduced grid efficiency as well as issues for the grid due to congestion and
transformer overloading.
The proliferation of these batteries has also created unprecedented opportunities for Pakistan’s
transition to a smarter electricity grid. Foremost amongst these opportunities is the potential to
improve energy efficiency and offer ancillary services to the electric grid by implementing smart
controllers. These smart controllers can operate with minimal communication requirements and can
already create substantial additional value for both householders and the grid operator. It is important
to note here that the capacity of installed batteries already outstrips the primary reserve in most
developing and developed countries.
 A number of challenges remain before any of these opportunities can be realized. These include
both regulatory and market-based issues. We recommend the setting up of energy (efficiency) labels
for such systems at the earliest to inform users not just about the upfront costs of such systems but
also their operational costs. Furthermore, proper dimensioning guidelines must be made available to
households to ensure minimum wastage, both in terms of investment and operational costs. Finally,
there is an urgent need to update the legal framework to enable demand side management pilots at
large scale. Recent innovations in trust-less systems (such as blockchain and peer-to-peer energy
trading can be relevant in this market segment).
Backup systems should however not be considered in isolation as this kind of thinking has
contributed to the creation of the problems faced today in the first place. A broader systems view is
required by relevant stakeholders in the government and industry. It is our belief that any potential
demand side management programs can benefit greatly from increased integration with local
production via solar PV and other heavy electric and thermal loads such as air conditioners. Updated
building and grid codes, price-based incentives and minimum legal requirements can help with the
widespread adoption of such measures.

References
Adenikinju, A. F. (2005). Analysis of the cost of infrastructure failures in a developing economy: The
case of the electricity sector in Nigeria.

Ahmad, A. S. (2016). Impact of wide-spread use of uninterruptible power supplies on Pakistan's

power system. Energy Policy 98, 629-636.

Amin, S. M. (2005). Toward a smart grid: power delivery for the 21st century. IEEE power and
energy magazine 3.5, pp. 34-41.
Andersson, G. D. (2005). Causes of the 2003 major grid blackouts in North America and Europe, and
recommended means to improve system dynamic performance. IEEE transactions on Power
Systems, 20(4), 1922-1928.
Antmann, P. (2009). Reducing technical and non-technical losses in the power sector.

Arshad, N. &. (2017). An analysis of the effects of residential uninterpretable power supply systems
on Pakistan's power sector. Energy for Sustainable Development, 36, 16-21.

Baosheng, T. (2003). Blackout in south of London and its lessons. POWER SYSTEM
TECHNOLOGY-BEIJING-, 27(11), 1-5.
Bayod-Rújula, A. A. (2009). Future development of the electricity systems with distributed
generation. Energy, 34(3), 377-383.

Boston, A. (2013). Delivering a secure electricity supply on a low carbon pathway. Energy Policy, 52,
55-59.

Brounen, D. &. (2011). On the economics of energy labels in the housing market. Journal of
Environmental Economics and Management, 62(2), 166-179.

Christiansen, W. &. (2006). Analysis of requirements in selected Grid Codes. Technical University of
Denmark (DTU).

Clement-Nyns, K. H. (2010). The impact of charging plug-in hybrid electric vehicles on a residential
distribution grid. IEEE Transactions on power systems, 25(1), 371-380.

Elia. (2013). Evolution ofancillary services needsto balancethe Belgian control area towards 2018.
Helm, D. (2002). Energy policy: security of supply, sustainability and competition. Energy policy,
30(3), 173-184.

Gram-Hanssen, K. B. (2007). Do homeowners use energy labels? A comparison between Denmark

and Belgium. Energy Policy, 35(5), 2879-2888.

IEA. (2011). Building code of Pakistan: energy provisions.

IEA. (2016). Energy policies of IEA countries: Belgium.

IEA. (2017). World Energy Outlook.

IRENA. (2017). Electricity storage costs.

IRENA. (2018). Pakistan: Renewables Readiness Assessment.

Kazmi, H. M. (2018). Gigawatt-hour scale savings on a budget of zero: Deep reinforcement learning
based optimal control of hot water systems. Energy, 144, 159-168.

Kessides, I. N. (2013). Chaos in power: Pakistan's electricity crisis. Energy policy, 55, 271-285.

Kundur, P. N. (1994). Power system stability and control. Vol. 7. McGraw-hill.

NEPRA. (2005). The grid code.

NEPRA. (2006). State of industry report .

NEPRA. (2016). State of the industry.

NEPRA. (2017). State of industry report.

Oudalov, A. C. (2007). Optimizing a battery energy storage system for primary frequency control.
IEEE Transactions on Power Systems, 22(3), 1259-1266.

Pasha, H. A. (1989). The economic cost of power outages in the industrial sector of Pakistan. Energy
Economics, 11(4), 301-318.

Pasha, H. A. (2013). The impact and cost of power load shedding to domestic consumers. The
Pakistan Development Review, 355-372.

PEC. (2014). Pakistan Engineering Council launches Building Energy Code of Pakistan. Retrieved from
[Link]
engineering-council-launches-building-energy-code-of-pakistan/

Schimschar, S. B. (2011). Germany's path towards nearly zero-energy buildings—Enabling the

greenhouse gas mitigation potential in the building stock. . Energy Policy, 39(6), 3346-3360.

Sohail, A. a. (2011). Energy Efficient Buildings in Pakistan. Science Vision.

Stevens, J. W. (1996). A study of lead-acid battery efficiency near top-of-charge and the impact on PV
system design. Twenty Fifth IEEE Photovoltaic Specialists Conference, (pp. 1485-1488).

Visscher, H. M. (2016). Improved governance for energy efficiency in housing. Building Research &
Information, 44(5-6), 552-561.

Warnock, V. J. (1986). Impact of voltage reduction on energy and demand: Phase II. IEEE
transactions on power systems, 1(2), 92-95.
World Bank. (2018). DataBank.

Young, K. W. (2013). Electric vehicle battery technologies. In Electric vehicle integration into modern
power networks (pp. 15-56). New York: Springer.