Energy 67 (2014) 1e18

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Assessment of the theoretical demand response potential in Europe

Hans Christian Gils*
Institute of Technical Thermodynamics, German Aerospace Center (DLR), Wankelstraße 5, 70563 Stuttgart, Germany

a r t i c l e i n f o a b s t r a c t

Article history: DR (Demand response) measures typically aim at an improved utilization of power plant and grid ca-
Received 6 March 2013 pacities. In energy systems mainly relying on photovoltaic and wind power, DR may furthermore
Received in revised form contribute to system stability and increase the renewable energy share. In this paper, an assessment of
3 February 2014
the theoretical DR potential in Europe is presented. Special attention is given to temporal availability and
Accepted 4 February 2014
geographic distribution of flexible loads. Based on industrial production and electricity consumption
Available online 1 March 2014
statistics, as well as periodic and temperature-dependent load profiles, possible load reduction and in-
crease is estimated for each hour of the year. The analysis identifies substantial DR potentials in all
Keywords:
Demand response
consumer sectors. They add up to a minimum load reduction of 61 GW and a minimum load increase of
Flexible electricity demands 68 GW, available in every hour of the year. The overall potential features significant variations during the
Demand side management year, which are characteristic for specific consumers and countries.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction and theoretical background
DR (Demand Response) actions are defined as “changes in
electric use by demand-side resources from their normal con-
sumption patterns in response to changes in the price of electricity,
or to incentive payments designed to induce lower electricity use at
times of high wholesale market prices or when system reliability is
jeopardized” [1]. In contrast to demand side management, which
also comprises energy efficiency measures and permanent and/or
regular utility-driven changes in the demand pattern, DR is focused
on load flexibility and short term costumer action [2,14]. In elec-
tricity systems primarily relying on fossil and nuclear fuels, in-
terventions in customer load can increase the profitability of power
plants and help to avoid investments in additional generation or
grid capacities [3,17]. Most demand side measures are thus
designed to either enable load curtailments in times of peak de-
mand or to shift loads to times of low demand [13]. In supply
systems with high capacities of intermittent solar PV(photovoltaic)
and wind power, the flexibility provided by DR can create addi-
tional benefits [4e6]. By adjusting the demand to the present
availability of fluctuating resources, curtailments can be reduced
and the overall RE (renewable energy) share can be increased. With
shifting times ranging from some minutes to a few days, DR com-
petes with alternative balancing options, such as batteries and
pumped storage power stations, as well as grid extension, but also
* Tel.: þ49 711 6862 477; fax: þ49 711 6862 747.
E-mail address: hans-christian.gils@dlr.de.
thermal storages in combined heat and power or heat pump sys-
tems [7e9]. Modifications in the demand pattern are typically
realized by direct or indirect load management programmes [10].
Existing DR measures include time-based rates on one hand, and
incentive based programmes on the other [11]. Technical require-
ment for the participation in DR programs is the availability of an
information and communication infrastructure allowing for the
transmission of and reaction to load, price and control signals.
Communication channels include radio and telecommunication, as
well as power lines [12,13]. Markets for flexible loads range from
on-site peak load reduction and increased internal PV consumption
to the participation in energy trade, the provision of operating
energy, as well as the clearance of imbalances in the TSO (trans-
mission system operator) area and the management of supply
shortfalls [14,15]. Due to various technical, economic, legal and
societal barriers, the use of DR is limited so far [16,17]. Existing
programmes are mostly focusing on large industrial consumers,
however, residential and commercial demand is increasingly also
taken into consideration [18].
Demand response resources have been identified in a broad
range of processes and devices [19,20,25,27,31,57]. In Germany and
Austria, their average shiftable and sheddable loads add up to
several GW [29,31,57]. The ENTSO-E (European Network of Trans-
mission System Operators for Electricity) has quantified a load
reduction potential of around 11 GW available throughout conti-
nental Europe [21]. Assuming a DR market potential equivalent of
2% of the annual peak load, a possible benefit of V53 billion ach-
ieved by smart meter installation and dynamic pricing on a Euro-
pean level has been estimated [22]. The impact of feedback and

http://dx.doi.org/10.1016/j.energy.2014.02.019
0360-5442/Ó 2014 Elsevier Ltd. All rights reserved.
2 H.C. Gils / Energy 67 (2014) 1e18
time-of-use tariffs on electricity demand and potential DR contri-
bution of residential appliances has been investigated in field trials
realized in Austria and Northern Italy [23,24].
A consumer and country-specific analysis of the flexible loads on
the European continent is missing so far. In this paper, the theo-
retical DR potential in Europe and North Africa is assessed.1 Its aims
are the characterization of electricity consumers that are able to
shift or shed their load for a given period of time and the provision of
a first estimate of their loads in Europe. Three different DR actions
are taken into account: load shedding, load shifting to an earlier time
and load shifting to a later time. Network load reductions achieved
by the usage of costumer-owned onsite generation are not included
in the analysis. DR potentials are determined across all demand
sectors: industry, as well as the tertiary and residential sector.
When assessing the potential future contribution of DR to the
system integration of RE, the temporal availability of flexible loads
is of particular importance [25]. The demand response behaviour of
non-residential consumers is directly correlated to industrial pro-
duction activity and business hours. In the residential sector, time-
related electricity demands can be derived from an evaluation of
household activity level and occupancy variance [26,27]. In the
work presented here, exemplary load profiles of all relevant con-
sumers are either estimated or extracted from metered data
available in literature. Based on these profiles, potentials for load
reduction and increase are calculated for each hour of the year. In
the context of the balancing of RE fluctuations, also the duration of
load interventions, the shifting time and the frequency of DR ac-
tions are of special importance. These parameters have decisive
impact on the quality of the corresponding DR potential. In order to
facilitate follow-on studies of the interaction of DR measures with
the electricity grid utilization, also the geographical distribution of
flexible consumers is evaluated. Given that the application of DR is
subject to constraints, different kinds of potentials need to be
treated separately. It can be distinguished between the theoretical,
technical, economic and practical potential [28,29]. Whereas the
theoretical potential comprises all facilities and devices of the
consumers suitable for DR, the technical potential includes only
those that can be controlled by the existing information and
communication infrastructure. A subset of the technical potential is
the economic potential of all DR consumers that can be operated in
a cost-efficient way. Another independent subset of the technical
potential arises from the acceptance of load interventions. The
effectively usable, practical potential consists of the intersection of
economic potential and accepted use. This work is restricted to the
assessment of the theoretical DR potential. Limitations for technical
reasons not related to industrial production processes, costs or
refusal to participate will at this point be neglected.
2. Data and methodology
The analysis is performed in four steps. First, processes and
appliances suitable for DR are identified (Section 2.1). Then, their
characteristic load profiles are assessed (2.2). In the third step,
annual electricity demands and installed capacities in the year 2010
are quantified, and a flexible load share for each consumer is
evaluated (2.3e2.5). Finally, the geographical distribution of DR
potentials is investigated (2.6).
1
Countries that are included: Albania, Algeria, Austria, Belarus, Belgium, Bosnia
and Herzegovina, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Egypt,
Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Libya,
Liechtenstein, Lithuania, Luxembourg, Macedonia, Malta, Moldova, Montenegro,
Morocco, Netherlands, Norway, Poland, Portugal, Romania, Serbia and Kosovo,
Slovakia, Slovenia, Spain, Sweden, Switzerland, Tunisia, Turkey, United Kingdom,
Ukraine.
2.1. Identification of shiftable loads and required parameters
In this study, a total of 30 different processes and appliances are
taken into consideration. Shiftable loads typically feature one of the
following characteristics: heat or cold storage (e.g. space heating,
refrigerators), demand flexibility (e.g. washing, ventilation) or
physical storages (e.g. cement industry, fresh water supply). In-
dustrial load shifting may be limited by technical constraints,
process requirements and availability of unutilized plant or ma-
chine capacity. For processes with very high utilization rates e as
they are found in energy-intensive industries e only load shedding
without previous or subsequent balancing can be implemented. In
residential and commercial sector, typically both load shifting and
shedding can be realized. Due to higher costs and losses of comfort,
this study evaluates load shedding only for energy-intensive in-
dustrial processes. Table 1 provides an overview of processes and
appliances included. The analysis is limited to those loads that can
be shifted or shedded for at least 1 h. Detailed descriptions of their
technical properties and DR behaviour can be gathered from the
references cited in Table 1.
For each country, DR consumer and hour of the year, potential
load increase Pincrease and load reduction Preduction are assessed.
Load reduction can be realized either by shedding or delaying de-
mand, whereas load increase is equivalent to advancing the oper-
ation of processes or devices. Loads can thus be shifted to both an
earlier or later point in time. In case of load shifting, every load
increase is followed by a decrease due and vice versa. Flexible loads
are calculated from characteristic load profiles and annual elec-
tricity demands. Furthermore, shares of used and free capacity that
can be activated for DR are assessed. Annual demands are obtained
from statistics or estimated based on industrial production capac-
ities and equipment rates of domestic appliances. Given that for the
North African countries no tertiary sector electricity demands and
appliance equipment rates are available, only potentials in energy
intensive industries are considered there.
Power demand flexibility can be compared to a functional en-
ergy storage with limited storage period. Its charging capacity is
determined by the flexible load, its reservoir capacity by the
maximum duration of DR interventions tinterfere, and its maximum
storage period by the shifting time tshift. The shifting time defines
the maximum duration until load that has been advanced or
delayed needs to be balanced again, whereas the intervention time
reflects a limit in duration of changes in the normal demand
pattern. Taking into account an annual limit in number of DR in-
terventions fDR, the storable energy per year can be calculated.
Parameters limiting DR are typically depending on process cycles,
physical storage capacities for intermediate products or the ther-
mal capacity of heated/cooled goods or rooms. The assumed values
for interference and shift times, as well as frequency of DR events
are summarized in Table 1.
2.2. Demand profiles of shiftable loads
In order to analyse the variability of the DR potential during the
year, exemplary load profiles are taken into account. As no own
measurements have been performed, metered data and informa-
tion about typical demand pattern available in literature are used.
The share of every hour in annual electricity demand sload,i is
evaluated separately for all consumers suitable for DR. Depending
on energy usage, the load profile is assumed to follow characteristic
periodic seasonal, weekly and daily profiles. For technologies
providing heat or cold, hourly demands are further correlated to
outside temperature. Whether the electricity demand is assumed
to depend on the time or temperature is indicated in Table 1 for
each DR consumer.
 H.C. Gils / Energy 67 (2014) 1e18 3

Table 1
Electricity consumers suited for DR with applicable measure, duration, shifting time, frequency, as well as temperature and time dependencies.

Process/Appliance DR measure tshift tinterfere fDR s(t) s(w) Reference

e h h 1/a e e

Industry Electrolytic primary aluminium Load shedding Na 4 40 No No [57]

Electrolytic refinement of copper Load shedding Na 4 40 No No [57]
Electrolytic production of zinc Load shedding Na 4 40 No No [57]
Steelmaking in electric arc furnaces Load shedding N 4 40 No No [57]
Chloralkali process (membrane/amalgam) Load shedding Na 4 40 No No [57]
Cement mills Advance/Delay 24 3 365 Season, Hour No [57]
Mechanical wood pulp production Advance/Delay 24 3 365 No No [57]
Recycling paper processing Advance/Delay 24 3 365 No No [57]
Paper machines Advance/Delay 24 3 365 No No [57]
Calcium carbide production Advance/Delay 24 3 365 No No [49]
Air liquefaction in cryogenic rectification Advance/Delay 24 3 365 No No [57]
Cooling in food manufacturing Advance/Delay 24 2 1095 Season, Hour No [57]
Ventilation w/o process relevance Delay 2 1 1095 Day No [57]
Tertiary Cooling in food retailing Delay 2 1 1095 Season, Hour No [31]
Cold storages Advance/Delay 2 2 1095 Season, Hour No [31]
Cooling in hotels and restaurants Advance/Delay 2 2 1095 Season, Hour No [31]
Commercial ventilation Delay 2 1 1095 Day, Hour No [31]
Commercial air conditioning Delay 2 1 1095 Hour Yes [31]
Commercial storage water heater Advance 12 12 1095 Hour Yes [31]
Commercial storage heater Advance 12 12 1095 Hour Yes [31]
Pumps in water supply Advance/Delay 2 2 1095 Hour No [63]
Waste water treatment Advance/Delay 2 2 1095 No No [62]
Residential Freezer/Refrigerator Delay 2 1 1095 Season, Hour No [29,31]
Washing machines, Tumble Drier, Dish washer Advance/Delay 6 Nb Nb Season, Day, Hour No [30,57]
Residential air conditioner Delay 2 1 1095 Hour Yes [76]
Residential electric storage water Advance 12 12 1095 Hour Yes [31]
Residential heat circulation pump Delay 2 1 1095 Hour Yes [31]
Residential electric storage heater Advance 12 12 1095 Hour Yes [31]
a
In the case of load shedding, the shifting time is infinite.
b
Given that in every hour different devices are switched on, there is no general limit in duration and frequency of the DR.

2.2.1. Constant and periodic seasonal, weekly and daily profiles
Energy-intensive production processes are typically running at
very high capacity utilization levels [46]. For this reason, a constant
load is applied during all hours of the year. Only exception is the
cement industry where utilization ranges between 40% and 100%.
In addition to winter times e when construction activities are
typically reduced e production is also lowered in the daytime on
workdays [32,46]. It is assumed that utilization in winter is by 20%
lower than in summer, and in the daytime on workdays at all
seasons reduced to two thirds of its night load. For industrial
ventilation energy demand, a weekend decline of 40% (Saturday)
and 50% (Sunday) is assumed; commercial ventilation is further-
more reduced by 50% at night-time. The electricity demand of
cooling appliances in private homes, retailing, hotels and restau-
rants is estimated to be by 10% lower in winter times than in
summer; it additionally declines by 20% at night, given that the
frequency of user interventions tends to go down. Based on
metered data presented in Ref. [33], cold storages and industrial
cooling are assumed to have lower demands on Saturdays (5%),
Sundays (10%) and during peak network load hours in the
morning (50%). Pumps in the fresh water supply are also typically
running during off-peak hours at night; here it is assumed that load
is reduced by two thirds in the daytime. The operation of washing
machines, tumble dryers and dish washers is mainly driven by the
daily routine of its users. Ref [35] provides measured hourly usage
profiles for different weekdays and seasons, which are used here.
days (HDD/CDD) are calculated for each country and, where avail-
able, Nuts-2 statistical region.
Daily shares Qd in annual space heating demand Qyear are
calculated with an extended degree day method, which considers
the HDD number nHDD,d not only of the current, but also the six
preceding days (see Equation (1)). In doing so, the HDDs in the
Nuts-2 regions of each country are weighted according to its pop-
ulation share. The hourly demand during each day is estimated
using temperature-dependent average profiles, which were
derived from the network load profile of a German district heating
supplier of residential and commercial consumers.
P 1
2a nHDD;da
a
Qd ¼ P 1 P365 $Qyear a ¼ 0.6 ðMWhÞ (1)
2a $ d ¼ 1 nHDD;d
a
With a difference of only 20% between coldest and warmest day
of the year, daily hot water demand is assumed to be almost con-
stant. This assumption is derived from the analysis of the same
district heating time series. The hourly demand within each day
follows the measured load profile presented in Ref. [35].
Daily shares in annual air conditioning demand are also calcu-
lated with Equation (1). For cooling demand, however, only the
CDD of the current and previous day are taken into consideration.
Within each day, air conditioning power demand is assumed to
peak in afternoon hours, in accordance with the representative
profile shown in Ref. [76].

2.2.2. Temperature-dependent profiles

The energy demand of space heating, hot water generation and
air conditioning is strongly correlated to outside temperature. Us-
ing a set of GIS maps containing European daily average tempera-
tures in the year 2006 at a spatial resolution of 7 km [34],
temperature profiles, as well as daily heating and cooling degree
2.3. Flexible loads in industry
2.3.1. Energy-intensive processes
DR potentials in energy-intensive industries are estimated
based on production capacities Ci and specific energy demands
4 H.C. Gils / Energy 67 (2014) 1e18

spec
Wi found in literature and statistics. Annual electricity demands Table 3
Wi and installed electrical capacities Pinstalled,i are calculated ac- Parameter for the calculation of DR potentials in industrial cross-sectional
technologies.
cording to Equations (2) and (3), taking into account the capacity
utilization level suse,i, total number of hours of the year Nhours and Process Wi nFLH sreduction sincrease References
revision outages srevision,i. Most industrial processes are operated at TWh/a h/a % %
utilization levels below 100%; the actual production is conse-
Cooling in food 31 5840 50% 90% [56e58,75]
quently lower than the maximum production capacity. It is manufacturing
assumed that the production capacity reflects the quantity that can Ventilation w/o 12 7008 50% 0% [56e58,71,75]
be manufactured if the unit is running at full load at all times except process relevance
for its annual revision.

spec
Wi ¼ Ci $Wi $suse;i ðMWhÞ
Wi
Pinstalled;i ¼   ðMWÞ
Nhours $suse;i $ 1  srevision;i
The potential load reduction Preduction,i (t) in each hour is given
by the difference between current load and minimum load of the
process (see Equation (4)). Its value changes during the year ac-
cording to the load profile sload,i (t). The minimum process load is
defined relative to the installed electrical capacity and given by
parameter sminimum,i. The potential load increase Pincrease,i (t) is
calculated from the difference between maximum load and current
load, which is at least temporarily greater than zero for all pro-
cesses operated at less than 100% utilization. This difference is
multiplied with parameter sincrease,i, reflecting the free production
capacity share available for DR (see Equation (5)). Table 2 sum-
marizes the assumed parameter values.
Preduction;i ðtÞ ¼ sload;i ðtÞ$Wi  Pinstalled;i $sminimum;i
|fflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
Load in hour t Minimum Load
(2) 2.3.2. Industrial cross-sectional technologies
Flexible loads in the cross-sectional technologies cooling and
ventilation are evaluated based on their annual electricity demand
(3) Wi. These demands are estimated using data from Refs. [56,58,71].
Dividing Wi by the number of full load hours nFLH, the installed
capacity is obtained (see Equation (6)). In contrast to energy-
intensive processes, no revision outage is considered.
Wi
Pinstalled;i ¼ ðMWÞ (6)
nFLH;i
In the assessment of potential load reduction and increase, fixed
shares in current load sreduction,i and unused capacity sincrease,i
available for DR are assumed. Taking into account the load in the
current and subsequent hours, they allow for the calculation of
shiftable loads according to Equations (7) and (8). In case of a load
increase, all demand of the following tshift hours can be advanced.
Its upper limit is set by the installed capacity. Estimated energy
demands, utilization levels and DR shares of the industrial cross-
ðMWÞ (4)
sectional technologies are summarized in Table 3.
Preduction;i ðt; wÞ ¼ Wi $sload;i ðt; wÞ $ sreduction;i ðMWÞ (7)
|fflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflffl{zfflfflfflfflfflfflffl}
Load in hour t Shiftable Share

   
Pincrease;i ðtÞ ¼ Pinstalled;i $ 1  srevision;i  sload;i ðtÞ$Wi $ sincrease;i ðMWÞ (5)
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflffl} |fflfflfflfflffl{zfflfflfflfflffl}
Maximum Load Load in hour t Shiftable Share

Table 2
Parameter for the calculation of DR potentials in energy-intensive industries.
spec
Process Ci Wi srevision suse sminimum sincrease References

Mt/a kWh/t % % % %

Aluminium electrolysis 4.4 14000 5% 100% 75% 0% [36,46,57,77]

Copper electrolysis 3.4 350 5% 95% 75% 0% [38,57,77]
Zinc electrolysis 2.2 3400 5% 100% 75% 0% [37,38,57,77]
Electric steel production 127 525 5% 100% 0% 0% [39,46,77]
Chloralkali process e membrane 6.5 2100 5% 95% 50% 0% [40,41,57]
Chloralkali process e mercury 4.1 3600 5% 95% 30% 0% [40,41,57]
Mechanical wood defibration 19 1500 5% 80% 0% 100% [44e46,57]
Wastepaper processing 75 250 5% 80% 0% 100% [42e45,57]
Paper machines 122 425 5% 90% 70% 100% [45,57,64]
Cement mills 550 110 5% 80% 50% 100% [46,47,77]
Calcium carbide production 0.5 3100 5% 80% 0% 100% [48,49]
Air separation e O2 23 238 5% 80% 60% 100% [50]
Air separation e N2 14 160 5% 80% 60% 100% [50]
Air separation e Ar 8 224 5% 80% 60% 100% [50]
 H.C. Gils / Energy 67 (2014) 1e18 5

0 1

X
tþt B
shift B
C
C!
Pincrease;i ðt; wÞ ¼ B Wi $sload;i ðt0 Þ $ sincrease;i C Pinstalled;i  Wi $sload;i ðt; wÞ ðMWÞ (8)
B C
t0 ¼ t @ |fflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflffl{zfflfflfflfflffl} A |fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
Load in hour t0 Shiftable Share Unused Capacity

2.4. Flexible loads in tertiary sector
Tertiary sector DR potentials are available in the supply of cold,
heat, water and ventilation, as well as in waste water treatment.
Flexible loads in these applications are calculated based on their
annual energy consumptions. In the absence of country-specific
data, they are approximated by multiplying the tertiary sector de-
mand [51e53] with average demand shares of the relevant uses.
According to survey data published in Ref. [72], 19.7% of the 2007
electricity consumption in EU-27 countries was used for the supply
of space heat and hot water, 12.6% for ventilation, 5.9% for pumps,
8.7% for cooling appliances and 2.9% for air conditioning. All other
uses are not relevant to DR. With exception of space heating and air
conditioning, which are assumed to depend on outside tempera-
ture, these shares are applied to all European countries. Pursuant to
values estimated for Germany in Ref. [56], the electricity demand of
cooling appliances is subdivided to food retailing (75%), cold stor-
ages (10%) and hotels/restaurants (15%).
Wi ¼ Wtertiary $sdemand;i ðMWhÞ
The air conditioning share in the sectoral demand is approxi-
mated using long-term average CDDs of each country [54]. It ranges
between 0.5% in northern countries to 12% on the Mediterranean
islands (see Table 4 and Table 7). These values rely on shares
available for a number of countries and the overall energy demand
of air conditioning.
Annual full load hours of hot water boilers and storage heaters
are estimated for each country based on long-term average HDDs
(see Table 7) [55]. It is assumed, that the usage is higher in colder
climates, as to the values in Table 5. In the assessment of air
conditioner and residential heat circulation pump full load hours, in
addition to HDDs and CDDs, demand profiles for heat and cold are
taken into account. Their calculation is described in Section 2.2.2. It
is assumed that whenever the cooling demand surpasses 60% of its
peak value, the overall air conditioner park is running at the
maximum load of 75% of its installed capacity. For lower demands,
capacity utilization reaches 1.25-times the ratio between current
and peak load. Resulting full load hours range between 136 in
Estonia and 991 in Cyprus, with a European average of 467 h (see
Table 7). In countries without CDD, it is set to 100 h/year.
Table 4
Air conditioning (AC) share in commercial electricity demand.
With the annual electricity demands and full load hours sum-
marized in Table 6, the installed capacity is calculated according to
Equation (6). No procedural limitations of load shifting are
considered in tertiary sector. Load shares available for DR are thus
set to 100% of current load and unused capacity, respectively. Based
on annual demand, installed capacity and hourly load profiles,
possible load reduction and increase are calculated according to
Equations (7) and (8), respectively.
2.5. Flexible loads in the residential sector
Residential load shifting is evaluated for heating, cooling, air
conditioning and washing equipment. The latter comprises
washing machines, tumble dryers and dish washers. In contrast to
the other sectors, residential DR loads are quantified in a bottom-up
approach. Household numbers and country-specific equipment
rates ri of relevant devices are obtained from Refs. [64e67,71,73], or
approximated using available data (see Table 7). Multiplying the
(9) unit number with the specific capacity and energy consumption,
country sums are calculated for each appliance type according to
Equations (10) and (11). Based on [68,72], an annual refrigerator
and freezer unit demand of 350 kWh is applied.
unit
Pinstalled;i ¼ NHH $ri $Pinstalled;i ðMWÞ (10)
Wi ¼ NHH $ri $Wiunit ðMWhÞ (11)
For washing equipment, average power demands during use, as
well as frequency and duration of use are taken into account (see
Table 8). Relying on measured consumption data from Refs. [69,70],
annual demands are calculated (Equation (12)).
Wiunit ¼ Pcycle;i $dcycle;i $Ncycle;i ðMWhÞ (12)
Wiunit ¼ Pinstalled;i
unit
$nFLH
i ðMWhÞ (13)
In the calculation of annual space heating, hot water generation
and air conditioning electricity consumption, unit capacities are
multiplied with estimated full load hours (Equation (13)). Average
capacities per dwelling are assumed with 1.65 kW for air condi-
tioners, 2 kW for electric storage water heaters, 100 W for heat
circulation pumps and 14 kW for electric storage heaters [72e76].

Annual number of CDD sdemand,AC Table 5

Assumed annual full load hours for storage water heater (WH) and storage heater
<100 0.5% (SH).
<200 1.0%
<300 2.0% Annual number of HDD nFLH,WH nFLH,SH
<400 3.0%
h/a h/a
<500 4.0%
<600 5.5% 1000 175 200
<700 7.0% 2000 200 350
<800 8.5% 3000 225 500
<1000 10.0% 4000 250 650
1000 12.0% >4000 275 800
6 H.C. Gils / Energy 67 (2014) 1e18

Table 6 including settlement areas, agricultural use, forest and water-

Parameter for the calculation of tertiary sector DR potentials. bodies. Here, only the categories continuous urban fabric, discon-
Process/device sdemand nFLH sreduction sincrease References tinuous urban fabric and industrial or commercial units are taken into
% h/a % %
account. The tertiary sector DR potential is equally distributed over
all grid cells of these classes. In countries not included in the CLC
Cooling in food retailing 6.5% 5840 100% 0% [56,57,72,75]
data, the allocation is done according to the population (see Section
Cold storages 0.9% 5000 100% 100% [56,57,72,75]
Cooling in hotels and 1.3% 5000 100% 100% [56,57,72,75] 2.6.3).
restaurants
Commercial ventilation 12.6% 4380 100% 0% [57,58,72]
Commercial air conditioning -a -a 100% 100% [72,75] 2.6.3. Residential sector
Commercial storage 1.5% -a 0% 100% [59,72] Residential DR potentials are allocated according to the popu-
water heater lation distribution. The population grid used is derived from
Commercial storage heater -b -b 0% 100% own
Eurostat statistics, a GIS data set prepared by the JRC (Joint Research
assumptions
Pumps in water supply 3% 4380 100% 100% [60,61] Centre) [79] and the Grump data of the Center for International Earth
Waste water treatment 3% 5694 20% 50% [62,63] Science Information Network at Columbia University [80]. Eurostat
a
see Table 7.
provides the population in each Nuts-3 region of EU countries,
b
due to data availability only considered in Germany (sdemand ¼ 2%, nFLH ¼ 650) Norway, Switzerland, Croatia, Turkey and Liechtenstein. Within the
and France (sdemand ¼ 5%, nFLH ¼ 500). regions, the population is allocated according to the JRC data. For
the rest of Europe, the Grump data is used.

Country-specific air conditioner, water heater and storage heater

full load hours are obtained as described in Section 2.4. In the
calculation of heat circulation pump full load hours, an approach
similar to that used for air conditioners is chosen. According to
[75,76], average annual operation time of heat circulation pumps in
Europe is around 5000e6000 h. This implies that not all pumps are
switched off in summer. Here, a base load of 25% of the installed
capacity is assumed for all countries. This base load is assigned to all
hours with a demand below 15% of the annual peak. For higher
demands, the load is assumed to be 1.67-times the ratio between
current and peak load. If this ratio is higher than 0.6, circulation
pumps are assumed to run at full load. It is assumed that all devices
are available for DR; sreduction,i and sincrease,i are thus set to 100%.
Hourly load increase and decrease are then obtained using Equa-
tions (7) and (8).
2.6. Geographic allocation of flexible loads
The spatial distribution of DR potentials is assessed using high
resolution GIS data and industrial production statistics. Population
density and land use data allows for the allocation of flexible loads
to grid cells of 0.0083 side length. At the equator, this corresponds
to approximately 1 km, in the investigation area the cell area ranges
from 0.27 to 0.74 km2.
2.6.1. Industry
In energy-intensive industries, an identification of individual
production plants is pursued. Geographic coordinates of facilities
are determined, allowing for a detailed spatial allocation of flexible
loads. Based on [77] and different industry associations, the pro-
duction sites of aluminium, electric steel, copper, zinc, chlorine,
calcium carbide and partially also cement industry are identified.
The exact assignment of production capacities to factories cannot in
all cases be extracted from statistics and are estimated where
necessary. The remaining industrial DR potentials are allocated
according to employment statistics of Eurostat, which are available
for dedicated sectors and on Nuts-3 statistical region scale.
2.6.2. Tertiary sector
Given the high number of tertiary sector consumers, a
geographic allocation cannot account for individual sites. Instead,
high resolution GIS data containing residential and commercial
areas is used. For the predominant part of European countries, CLC
(Corine Land Cover) provides land use data in a spatial resolution of
100 m [78]. It assigns each grid cell to one of 44 land use classes
3. Results
3.1. Shiftable loads
Relying on the methodology and data introduced in chapter 2,
substantial amounts of flexible loads are identified. Aggregated
over all countries and consumers, the hourly average load reduc-
tion potential through shedding and delaying adds up to 93 GW.
With the assumed hourly load profiles, in the course of the year its
value varies from 61 GW to 172 GW. The potential shows similar
orders of magnitude in all demand sectors. Annual averages are
around 25 GW in industry, 31 GW in tertiary sector and 37 GW in
residential sector. In industry, the reduction potential is almost
constant throughout the whole year, whereas in both residential
and tertiary sector it varies from less than 20 GW to more than
75 GW.
The overall load increase that can be achieved by advancing
demands fluctuates between 68 GW and 499 GW, with an average
of 247 GW. This much higher potential results from the assumption
that for shifting times tshift longer than 1 h the demand of various
subsequent hours can be concentrated to only 1 h. Precondition is
that the corresponding devices would regularly be run within the
following tshift hours and that sufficient free capacity is available.
This is particularly important in the residential sector, where by
advancing the operation of washing equipment, boilers and storage
heaters the load can be augmented by up to 449 GW. If load is
advanced by more than 1 h, the DR potential in the following hours
is reduced, as flexible loads of subsequent hours are correlated by
the shifting time. The residential minimum load increase during a
year is 43 GW, the average 209 GW. In comparison, the potential
load increase is much lower in the other demand sectors; in tertiary
sector it is found to vary between 21 GW and 45 GW, in industry
between 2 GW and 8 GW, with average values of 33 GW and 5 GW,
respectively.
The share each demand sector holds in the yearly average of
potential load reduction and increase shows significant differences
between countries (see Fig. 1 and Fig. 2). The residential share in
reducible load ranges from 11% in Luxembourg to 62% in Albania,
whereas the tertiary share varies from 13% in Moldova to 45% in
Ireland and the industrial share from 3% in Malta to 69% in
Luxembourg. Considering load increase, a strong dominance of
residential and commercial sector can be observed, with shares
between 53% in Turkey and 97% in Moldova. With the assumptions
made here, industry load increases are very limited. This is reflected
by the resulting industrial share in DR loads, which amount to
 H.C. Gils / Energy 67 (2014) 1e18 7

Table 7
Country input for DR potentials in residential and tertiary sectora.

Country Wtertiary HDD CDD nFLH,AC nFLH,CP NHH rRF rFR rWM rTD rDW rAC rWH rCP rSH

TWh/a 1/a 1/a h/a h/a Mio. % % % % % % % % %

Albania 0.51 1724 683 440 4239 1.1 30% 90% 60% 5% 5% 4% 20% 35% 3%
Austria 12.32 3469 173 237 5072 3.7 85% 98% 94% 34% 70% 1% 35% 60% 5%
Belarus 3.97 4299 88 211 4311 3.8 30% 90% 50% 2% 5% 0% 20% 45% 5%
Belgium 22.18 2771 102 232 4846 4.7 62% 114% 89% 60% 55% 2% 26% 60% 5%
Bosnia-Herzegovina 1.44 2949 261 325 4486 1.3 30% 90% 60% 5% 5% 4% 20% 35% 3%
Bulgaria 8.1 2648 430 361 4088 2.9 55% 97% 80% 0% 4% 14% 20% 45% 5%
Croatia 5.27 2561 418 440 3927 1.6 67% 98% 92% 6% 20% 28% 20% 35% 3%
Cyprus 2.26 762 1091 991 3038 0.3 40% 100% 95% 10% 30% 60% 25% 35% 5%
Czech Republic 13.99 3517 108 262 4516 4.4 40% 90% 80% 20% 20% 2% 23% 60% 5%
Denmark 10.71 3438 40 203 5066 2.6 79% 147% 91% 63% 70% 0% 21% 55% 5%
Estonia 2.53 4393 38 136 4466 0.6 40% 99% 89% 20% 10% 2% 25% 60% 5%
Finland 17.83 5252 48 100 4999 2.6 60% 97% 95% 20% 65% 0% 46% 55% 20%
France 145.44 2475 241 336 4785 27.4 86% 100% 94% 31% 49% 5% 44% 60% 5%
Germany 136.17 3178 122 275 4758 39.8 68% 113% 86% 39% 64% 2% 11% 60% 4%
Greece 18.0 1531 923 626 3666 4.2 12% 115% 93% 6% 40% 70% 75% 35% 5%
Hungary 11.36 2888 256 419 3882 4.3 67% 88% 99% 1% 9% 3% 35% 60% 5%
Ireland 9.16 2876 19 100 6236 1.6 35% 100% 95% 62% 50% 0% 9% 60% 8%
Italy 85.62 2120 600 584 4477 27.5 34% 99% 97% 10% 45% 25% 34% 35% 5%
Latvia 2.42 4220 58 153 4335 0.9 7% 96% 83% 2% 5% 1% 40% 60% 5%
Liechtenstein 0.06 3207 137 148 5726 0.03 60% 100% 95% 40% 50% 2% 15% 60% 5%
Lithuania 2.83 4048 68 193 4183 1.3 25% 95% 85% 20% 20% 2% 32% 60% 5%
Luxembourg 1.96 3164 99 222 5089 0.2 60% 100% 95% 40% 50% 2% 15% 60% 5%
Macedonia 1.50 2647 430 344 4504 0.7 30% 90% 60% 5% 5% 4% 20% 35% 3%
Malta 0.63 543 1043 928 2695 0.2 66% 100% 97% 45% 43% 60% 25% 35% 5%
Moldova 0.86 3317 325 407 3705 1.3 30% 90% 50% 2% 5% 0% 20% 45% 5%
Montenegro 0.41 2949 261 293 4993 0.2 30% 90% 60% 5% 5% 4% 20% 35% 3%
Netherlands 34.27 2851 68 267 4814 7.6 55% 98% 99% 68% 55% 4% 9% 60% 5%
Norway 28.77 5202 43 1692 6320 2.2 93% 98% 92% 46% 75% 0% 18% 55% 30%
Poland 43.26 3562 100 304 4064 14.2 40% 90% 80% 20% 20% 2% 29% 60% 5%
Portugal 16.40 1152 345 467 3997 4.1 67% 100% 96% 24% 40% 3% 18% 35% 5%
Romania 7.58 3040 290 403 4134 7.9 30% 95% 80% 2% 5% 1% 20% 45% 5%
Serbia and Kosovo 6.18 2949 261 401 4155 3.2 30% 90% 60% 5% 5% 4% 20% 35% 3%
Slovakia 8.01 3305 158 319 4455 2.7 38% 102% 69% 20% 20% 1% 23% 60% 5%
Slovenia 3.09 3024 189 341 4523 0.8 81% 98% 98% 45% 51% 15% 50% 60% 5%
Spain 88.05 1784 702 614 4107 17.1 45% 100% 99% 5% 45% 45% 36% 35% 5%
Sweden 32.75 4630 45 186 5346 4.7 98% 100% 76% 50% 68% 0% 14% 55% 20%
Switzerland 17.72 3411 137 243 5910 3.6 60% 100% 95% 40% 50% 2% 5% 60% 5%
Turkey 45.11 2520 641 548 4296 16.2 15% 50% 40% 2% 2% 15% 5% 10% 0%
UK 97.21 2954 66 148 5657 27.1 46% 106% 96% 58% 39% 0% 24% 60% 8%
Ukraine 24.60 3752 224 456 3996 16.7 30% 90% 50% 2% 5% 0% 20% 45% 5%
a
AC e air conditioning, CP e heat circulation pump, RF e refrigerator, FR e freezer, WM e washing machine, TD e tumble dryer, DW e dish washer, WH e storage water
heater, SH e storage space heater.

values between 1% and 5% in most countries. The remaining load residential sector, almost half of the load reduction can be realized
increase of 3%e39% is provided by tertiary sector. by shifting electricity consumption of freezers and refrigerators.
Flexible loads are distributed very unevenly over the 30 pro- Washing machines and heat circulation pumps provide around one
cesses and appliances analysed. Taking into account annual aver- fifth each; lower but still substantial potentials are found for dish
ages, highest reducible loads are found in pulp and paper (7%), steel washers, air conditioning and tumble dryers.
(9%) and cement (6%) industry, as well as commercial ventilation In contrast to the diversified distribution of reducible loads, the
(15%) and refrigerators/freezers in retailing (8%) and private potential load increase can almost completely be attributed to
households (17%). The industrial potential furthermore includes electric space heating (45%), storage water boilers (25%) and
considerable loads of aluminium electrolysis, Chloralkali process washing equipment (27%). This is a consequence of the assumed
and cross-sectional technologies. Minor load reductions can be technical restrictions in the industrial and commercial sector con-
made available in the copper, zinc, calcium carbide and air lique- cerning possibilities to advance loads. Average, minimum and
faction industry. The potential load turn down in tertiary sector maximum load reduction and increase potentials of the dominating
load is dominated by cooling appliances and fans, with smaller technologies are displayed in Figs. 3 and 4, respectively. Country
contributions from public water supply and treatment. In averages of all technologies are listed in Tables 9 and 10.
The subdivision of DR potentials to the investigated countries is
shown for each sector in Fig. 5. The comparatively high share non-
Table 8 EU countries hold in industrial load reduction results from the
Parameter for the calculation of residential sector DR potentials. additional consideration of the five Mediterranean countries of
Northern Africa. Industrial potentials are particularly high in
Device Wiunit Pcycle, i dcycle, i Ncycle, i References
Finland, Sweden and Luxembourg, whereas they are comparatively
kWh/a kW h 1/a
low in Denmark, the Netherlands and the United Kingdom.
Washing machines 219 0.75 2 146 [70,71] In order to assess the contribution of DR to power system
Laundry drier 307 1.5 2 102 [70,71] balancing, potential load reduction and increase are related to the
Dish washer 270 0.65 2 208 [69e71]
annual peak load in each country. Fig. 6 shows the ratios of
8 H.C. Gils / Energy 67 (2014) 1e18

100%
80%
60%
40%
Industry
20% TerƟary
ResidenƟal
0%

Greece

Ireland
Italy

Portugal

Switzerland
Bosnia & Herzegovina

Turkey
Austria

Germany
Belarus
Belgium

Bulgaria

Denmark

Hungary

Latvia

Malta
Moldova
Montenegro
Netherlands
Norway
Albania

Cyprus

Estonia

Romania
Serbia & Kosovo
Finland

Slovakia

United Kingdom
France

Macedonia
Liechtenstein
Lithuania

Poland

Slovenia

Sweden
Spain

Ukraine
CroaƟa

Czech Republic

Luxembourg
Fig. 1. Sectoral shares in average potential load reduction by country.

Fig. 2. Sectoral shares in average potential load increase by country.

minimum, maximum and average load reduction potential to the results from the assumed load concentration of residential appli-
2010 peak load. In most EU countries, the average reduction equals ances, which would usually be used in a later moment. The ratio of
between 10% and 20% of peak load. Lower values than that are average load increase potential and peak load shows significant
found in the Czech Republic, Macedonia, Norway, Serbia and difference between countries and ranges from 4% in Slovenia to 65%
Slovenia, higher values in Greece and Luxembourg. The ratio of in Lithuania.
maximum load reduction potential to peak load reaches very high
values in the Mediterranean countries, especially in Cyprus, Greece 3.2. Temporal availability of flexible loads
and Spain, where it reaches more than 50%. This arises from the
widespread use of air conditioning systems, which are assumed to The substantial difference between minimum and maximum
be available for load shifting. values of some flexible loads displayed in Figs. 3 and 4 indicate a
In a number of countries the potential load increase exceeds the strong temporal variation in the availability of DR potentials. They
2010 peak load in at least 1 h of the year (see Fig. 7). This again are particularly pronounced for space heating, ventilation and air

35000
MW
30000 Load ReducƟon Min 46963
25000
Load ReducƟon Max
20000
Load ReducƟon Average
15000
10000
5000
0

Fig. 3. Potential load reduction, subdivided by consumer.

 H.C. Gils / Energy 67 (2014) 1e18
140000
MW
120000 Load Increase Min
100000
Load Increase Max
80000
60000 Load Increase Average
40000
20000
0
Fig. 4. Technology shares in average potential load increase, subdivided by consumer.
conditioning, as well as residential washing equipment. Fig. 8 il-
lustrates the development of the daily load reduction average
during one year for five representative technologies. It reflects the
load profiles assumed in Section 2.2. With no load changes
considered, DR potentials in energy-intensive industries e here the
aluminium electrolysis is selected as example e are constantly
available throughout the whole year. Also the power demand of
food cooling appliances in all sectors shows only minor variations.
In contrast, shiftable loads in the provision of air conditioning and
space heating are strongly influenced by outside temperature and
have annual load curves contrasting each other. Particularly in air
conditioning demand, where short-term reactions to temperature
occur, high peaks in single days and hours can be observed. The
variations in the profile of the load reduction potential of com-
mercial ventilation are directly correlated to the assumed weekend
demand decline. Due to the high air conditioner share in overall
theoretical load reduction potential, the annual maximum is
reached in summer times. In peak hours, the potential can be twice
as high as on average. Hours with lowest reduction potential are
found in the transition period between winter and summer, when
both air conditioner and heat circulation use is low. Average daily
reduction potentials are found to be lower on weekends than on
working days. This results from reduced industrial and commercial
demand.
The flexible load is varying considerably also within each day.
This is illustrated for a selection of consumers and a representative
spring week in Fig. 9. Again, the DR potential provided by industry
and cooling appliances is fairly constant throughout the day. The
same applies to residential heat circulation pumps. In contrast to
that, the DR availability of washing equipment, air conditioners and
fans is heavily fluctuating. Due to the typical utilization cycles
driven by daylight, working hours and temperature, potentials are
mostly available during daytime. This coincides with the overall
system load pattern.
The identified load increase is dominated by washing and
heating equipment. Its power demands are fluctuating due to
different usage intensity and dependency on outside temperature.
With a shifting duration of 12 h and typical night-time peaks, the
potential load increase of storage heaters is highest in evening
hours. Using the demand profiles introduced in Section 2.2, the
overall load increase potential is higher in winter than in summer,
on weekends than on working days and at noon than during night-
time.
The temporal availability of load flexibility arises from the
composition of the overall DR potential. Differences between
countries are primarily associated to the fraction of space heating
and air conditioning. Fig. 10 and Fig. 11 include profiles for load
9
202297
reduction and load increase, respectively, in five representative
European countries, all normalized to the corresponding maximum
potential. In Finland, where the industrial share in the potential is
very high and almost no air conditioning is used, daily averages of
load reduction remain almost constant during the year. In contrast,
strong summer peaks can be observed in Spain and Italy. Due to a
more frequent use of washing and heating equipment in winter
times, the potential load increase is relatively higher in winter. This
tendency is at hand in all countries considered here; its exact shape
depends on the installed capacity of electric space and water
heating systems, as well as the course of the heating season.
3.3. Geographic allocation of flexible loads
DR potentials are concentrated to centres of population and
energy intensive industry production. Fig. 12 and Fig. 13 show the
potential load reduction and increase density in each grid cell.
Major cities can be easily identified there.
For further analysis of the spatial allocation of flexible loads,
sums over each Nuts-3 statistical region are formed, and average
values per km2 and inhabitant are calculated. The regional density
of the load reduction potential is displayed in Fig. 14. It reaches high
values not only in densely populated regions, but also those with a
concentration of energy intensive energies. The highest values of
more than 900 kW/km2 are found in Paris, Inner London and the
industrial city of Ludwigshafen am Rhein in Germany. Compara-
tively low densities are present in sparsely populated areas, for
example in north-eastern Germany, Scotland or northern Finland,
Norway and Sweden.
Taking into account population density, regions with high in-
dustrial and commercial DR potentials can be identified. In Fig. 15,
the per capita load reduction is shown for each region. Compara-
tively high values are found for example in the French region of
Aquitaine, the Norwegian coast and Luxembourg. For most Non-EU
countries, no regional data was available. The figure consequently
provides national average values. They are particularly low in the
North African countries, where only potentials in the energy-
intensive industries have been considered. Given that most in-
crease potential through advancing load is provided by residential
appliances, the geographic distribution is very similar to the pop-
ulation density. Inhabitant-specific values are almost homogenous
within all countries.
4. Discussion
The analysis reveals a substantial theoretical DR potential in
Europe. By shedding or delaying, an average load reduction of
 10
Table 9
Average potentials for load reduction by shedding or shifting to a later point in time, subdivided by country and consumer in MW.

Country/processa 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 21 22 23 24 25 26 27 30

Albania 0 0 0 0 0 0 0 15 21 0 0 0 0 4 1 1 7 4 2 0 52 16 2 2 4 19
Algeria 0 0 2 0 0 0 1 12 189 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Austria 0 1 0 7 67 46 44 46 57 2 4 27 14 91 12 18 178 14 42 8 268 86 44 79 1 124
Belarus 0 0 0 0 0 3 0 156 35 0 0 0 0 29 4 6 57 2 14 3 184 48 3 6 0 86
Belgium 0 3 22 131 28 17 65 153 103 0 8 68 16 164 22 33 320 25 76 15 333 105 100 80 4 160
Bosnia-Herzegov. 44 0 0 0 0 1 0 32 12 0 0 0 0 11 1 2 21 3 5 1 63 20 2 2 3 24
Bulgaria 0 1 5 13 0 4 2 84 64 0 0 20 5 60 8 12 117 37 27 5 176 58 0 3 27 64
Croatia 0 0 0 0 8 5 0 35 49 0 0 10 3 39 5 8 76 24 18 4 106 37 3 10 37 28
Cyprus 0 0 0 0 0 0 1 0 18 0 0 3 0 17 2 3 33 31 8 2 16 7 1 3 32 4
Czech Republic 0 0 0 30 15 9 13 64 66 0 1 25 24 104 14 21 202 16 48 10 228 88 31 27 3 139
Denmark 0 0 0 0 0 4 34 0 27 0 3 35 8 79 11 16 154 6 37 7 239 60 58 57 0 85
Egypt 108 0 0 0 0 4 11 360 511 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Estonia 0 0 0 0 21 1 2 0 10 0 0 5 2 19 3 4 37 1 9 2 34 14 4 2 0 18
Finland 0 1 21 14 797 126 21 168 12 0 2 30 11 132 18 26 257 10 61 12 160 61 18 51 0 78
France 135 0 18 198 1000 87 170 409 272 0 24 302 91 1078 144 216 2095 332 498 100 2032 644 293 418 87 922
Germany 211 5 12 575 249 204 449 931 422 12 51 275 147 779 104 156 1962 249 466 93 2885 854 543 791 32 1330
Greece 56 0 0 9 0 4 5 432 200 0 2 34 9 133 18 27 259 205 62 12 213 98 9 52 346 69

H.C. Gils / Energy 67 (2014) 1e18

Hungary 0 0 0 37 0 5 12 30 46 0 1 19 6 84 11 17 164 26 39 8 269 108 2 12 10 129
Ireland 0 0 0 1 0 0 13 0 20 0 15 25 6 68 9 14 132 5 32 6 86 38 35 25 0 64
Italy 65 1 18 41 57 89 159 1181 540 0 17 198 94 635 85 127 1233 684 293 59 1461 667 96 378 756 514
Latvia 0 0 0 0 0 1 2 0 4 0 0 5 1 18 2 4 35 1 8 2 37 19 0 1 0 27
Libya 0 0 0 0 0 0 0 75 97 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Liechtenstein 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 1
Lithuania 0 0 0 0 0 1 3 0 14 0 0 0 0 21 3 4 41 2 9 2 61 27 9 8 1 37
Luxembourg 0 0 0 0 0 0 2 171 15 0 0 2 4 15 2 3 28 1 7 1 14 5 3 3 0 7
Macedonia 0 0 0 0 0 0 0 22 10 0 0 0 0 11 1 2 22 7 5 1 34 11 1 1 2 13
Malta 0 0 0 0 0 0 0 0 0 0 0 1 1 5 1 1 9 9 2 0 11 4 3 2 17 2
Moldova 0 0 0 0 0 0 0 66 11 0 0 0 0 6 1 1 12 3 3 1 64 17 1 2 0 28
Montenegro 40 0 0 0 0 0 0 12 0 0 0 0 0 3 0 1 6 1 1 0 10 3 0 0 0 4
Morocco 0 0 0 0 0 1 1 31 132 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Netherlands 34 0 21 84 19 28 66 10 62 0 12 105 19 254 34 51 494 20 118 24 462 186 180 128 14 256
Norway 396 0 12 32 252 18 13 36 22 0 1 44 9 213 28 43 414 16 98 20 166 50 35 50 0 81
Poland 17 3 6 48 11 26 44 270 198 0 3 78 24 321 43 64 623 49 148 30 737 284 99 88 12 412
Portugal 0 0 0 10 0 14 21 108 121 0 1 27 12 122 16 24 236 56 56 11 272 98 35 50 10 72
Romania 91 0 0 50 5 5 8 255 223 0 0 26 14 56 7 11 109 17 26 5 397 159 6 12 6 175
Serbia and Kosovo 0 1 5 0 0 2 0 0 48 0 0 0 0 46 6 9 89 14 22 4 156 49 6 5 10 57
Slovakia 56 0 0 12 0 8 7 36 47 7 0 8 9 59 8 12 115 9 27 5 152 47 19 17 1 86
Slovenia 29 0 0 2 9 7 4 42 17 0 1 4 4 23 3 5 44 4 10 2 54 19 12 12 7 24
Spain 135 3 32 111 16 59 162 875 529 2 11 188 60 653 87 131 1268 854 301 60 989 422 30 237 888 304
Sweden 37 2 0 19 629 105 46 108 34 3 8 39 19 243 32 49 472 19 112 22 373 90 82 99 0 157
Switzerland 0 0 0 4 23 14 38 79 50 0 0 0 0 131 18 26 255 20 61 12 231 86 50 56 3 139
Tunisia 0 0 0 0 0 1 0 12 77 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Turkey 22 0 2 0 2 14 29 1375 860 2 0 0 0 334 45 67 650 360 154 31 420 162 11 10 251 84
UK 14 0 0 84 45 46 246 159 150 0 24 180 77 721 96 144 1400 55 333 67 1645 649 549 326 2 1016
Ukraine 0 0 1 0 0 8 10 98 149 0 0 0 0 182 24 36 354 56 84 17 802 209 12 26 4 361
a
Legend: 1-Aluminium; 2-Copper; 3-Zinc; 4-Chlorine; 5-Pulp; 6-Paper; 7-Recycling paper; 8-Steel; 9-Cement; 10-Calcium carbide; 11-Air separation; 12-Industrial cooling; 13-Industrial ventilation; 14-Cooling retailing; 15-
Cold storages; 16-Cooling hotels/restaurants; 17-Commercial ventilation; 18-Commercial AC; 19-Commercial storage water; 20-Commercial storage heating; 21-Water supply; 22-Water treatment; 23-Freezers/refrigerators;
24-Washing machines; 25-Tumble dryers; 26-Dish washers; 27-Residential AC; 28 Residential storage water; 29-Residential storage heater; 30-Heat circulation pump.
 H.C. Gils / Energy 67 (2014) 1e18 11

Table 10
Average potentials for load increase by shifting to an earlier point in time, subdivided by country and consumer in MW.

Country/Process2 5 6 7 8 9 10 11 12 15 16 19 20 21 22 24 25 26 28 29

Albania 0 0 0 0 5 0 0 0 0 1 10 0 2 1 98 11 10 119 181

Algeria 0 0 0 0 47 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Austria 17 28 11 0 14 3 5 25.3 9 14 253 0 34 11 517 263 473 877 1416
Belarus 0 2 0 0 9 0 0 0.0 3 4 82 0 11 4 289 16 36 580 1432
Belgium 7 10 16 0 26 0 10 63.5 16 25 456 0 62 21 631 598 479 759 1556
Bosnia-Herzegov. 0 1 0 0 3 0 0 0.0 1 2 30 0 4 1 118 14 12 162 263
Bulgaria 0 2 1 0 16 0 0 18.3 6 9 166 0 22 7 348 1 20 357 922
Croatia 2 3 0 0 12 0 0 9.3 4 6 108 0 15 5 223 20 60 199 294
Cyprus 0 0 0 0 5 0 0 2.8 2 3 46 0 6 2 41 6 16 34 49
Czech Republic 4 6 3 0 17 0 2 23.4 10 16 287 0 39 13 526 184 162 675 1554
Denmark 0 2 9 0 7 0 4 32 8 12 220 0 30 10 361 348 342 380 967
Egypt 0 2 3 0 128 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Estonia 5 0 0 0 3 0 0 4 2 3 52 0 7 2 81 26 11 115 236
Finland 199 77 5 0 3 0 2 27 13 20 366 0 49 16 364 107 307 885 4121
France 250 53 42 0 68 0 28 280 108 162 2988 6711 402 134 3869 1757 2500 7424 8842
Germany 62 125 112 0 105 14 61 255 78 117 1354 2122 376 125 5127 3261 4734 2999 11292
Greece 0 2 1 0 50 0 2 32 13 20 370 0 50 17 585 52 312 1721 1079
Hungary 0 3 3 0 12 0 1 18 8 13 233 0 31 10 646 9 72 938 1273
Ireland 0 0 3 0 5 0 18 23 7 10 189 0 26 9 229 207 148 89 986
Italy 14 55 40 0 135 0 20 183 63 95 1759 0 237 79 4006 578 2260 5677 8761
Latvia 0 0 1 0 1 0 0 5 2 3 50 0 6 2 112 3 7 270 345
Libya 0 0 0 0 24 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Liechtenstein 0 0 0 0 0 0 0 0 0 0 1 0 0 0 2 1 1 1 6
Lithuania 0 1 1 0 3 0 0 0 2 3 58 0 7 2 161 53 47 305 468
Luxembourg 0 0 0 0 4 0 0 2 1 2 40 0 6 2 30 18 20 22 78
Macedonia 0 0 0 0 3 0 0 0 1 2 31 0 4 1 63 7 6 86 140
Malta 0 0 0 0 0 0 0 1 0 1 13 0 2 1 23 15 13 19 23
Moldova 0 0 0 0 3 0 0 0 1 1 18 0 3 1 100 6 12 182 426
Montenegro 0 0 0 0 0 0 0 0 0 0 8 0 1 0 19 2 2 27 45
Morocco 0 1 0 0 33 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Netherlands 5 17 16 0 16 0 14 98 25 38 704 0 95 32 1116 1079 767 419 2474
Norway 63 11 3 0 5 0 1 41 21 32 591 0 79 26 300 210 301 294 5937
Poland 3 16 11 0 50 0 4 72 32 48 889 0 120 40 1704 596 525 2771 4878
Portugal 0 9 5 0 30 0 1 25 12 18 337 0 45 15 589 210 300 392 1043
Romania 1 3 2 0 56 0 1 24 6 8 156 0 21 7 953 33 73 1088 2749
Serbia and Kosovo 0 1 0 0 12 0 0 0 5 7 127 0 18 6 292 34 30 400 618
Slovakia 0 5 2 0 12 8 0 8 6 9 164 0 22 7 282 115 101 430 958
Slovenia 2 4 1 0 4 0 1 4 2 3 63 0 8 3 111 72 71 260 275
Spain 4 36 41 0 132 3 13 174 65 98 1809 0 243 81 2535 179 1419 3366 4449
Sweden 157 64 11 0 9 4 9 36 24 36 673 0 90 30 537 495 592 497 7516
Switzerland 6 8 9 0 13 0 0 0 13 20 364 0 49 16 514 303 333 124 1480
Tunisia 0 1 0 0 19 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Turkey 1 9 7 0 215 2 0 0 33 50 927 0 124 41 972 68 60 499 0
UK 11 28 61 0 37 0 29 167 72 108 1997 0 269 90 3898 3297 1950 4003 15463
Ukraine 0 5 2 0 37 0 0 0 18 27 505 0 68 23 1256 70 155 2293 5634

93 GW can be achieved. This average value is equivalent to one
third of the 280 GW minimum load, and approximately one sev-
enth of the 620 GW peak load measured in the European part of the
investigation area in 2010. The average load increase through
bringing forward electricity demand adds up to 247 GW. This very
high value results from the assumption that demand of various
hours can be accumulated to only 1 h. The minimum DR potentials
available in every hour of the year amount to 61 GW of load
reduction and 68 GW of load increase.
In the evaluation of the calculated theoretical potentials, it
needs to be taken into account that both load and DR potential of
subsequent hours are influenced, whenever a load reduction or
increase is called. All shifted demand needs to be balanced within a
given period of time, thus increasing or decreasing the load in one
or several following hours. Given that one and the same load cannot
be brought forward twice, the load increase potential of different
hours is correlated to each other. Due to limitations in duration and
frequency of DR interventions, this is also the case for the load
reduction potential.
The load shifting potential is dominated by customers allowing
only short interventions and shifting times (see Table 1). They
include cooling processes in all sectors, air conditioning,
ventilation, heat circulation pumps, fresh water supply and
wastewater treatment. All these appliances can be interrupted for
only one or 2 h, and the reduced load typically needs to be recov-
ered within the same time span. This limits the storage period of
the functional storage provided by DR. Consequently, the field of
application of these DR consumers in energy systems with high
renewable shares is restricted to the balancing of short term fluc-
tuations. Depending on the hour of the year, the energy that within
2 h can be charged into the virtual storage by load reduction ranges
from 47 GWh to 141 GWh. The temporal variation in storage ca-
pacity is correlated to the assumed load profiles of consumers
participating in DR. The discharging has to begin immediately af-
terwards and is also limited to a duration of 2 h. The amount of
energy that can be stored by advancing the short-term DR loads is
smaller, and varies from 4 GWh to 12 GWh.
Longer intervention times are found for washing equipment, as
well as residential and commercial electric space and water
heaters. With the assumed charging durations of up to 12 h for
electric heat production and 6 h for washing equipment, the energy
storage capacity is comparatively high. Due to its dependency on
outside temperature and appliance usage pattern, it shows strong
variations between 75 GWh and 567 GWh. Load delay is only
12 H.C. Gils / Energy 67 (2014) 1e18

Fig. 5. Country shares in average load reduction (left) and increase (right), subdivided by sector.

Fig. 6. Share of minimum, maximum and average load reduction potential in yearly peak load, subdivided by country.

considered for washing equipment. Its storage capacity can reach
up to 217 GWh during daytime and goes down to zero at night. The
much better storage function of washing and heating equipment
comes along with shortcomings concerning efficiency and accep-
tance. Due to high exergy losses, the use of electricity for the
resistive generation of heat is unfavourable compared to alterna-
tives, such as district heating or condensing boilers. Even though
the use of resistive electric air and water heaters is likely to be
reduced in the coming decades in the framework of energy effi-
ciency efforts, existing units can deliver an important contribution
to DR programs. In contrast to electric storage space and water
heaters, shifting of washing equipment entails significant con-
straints for users and might be accepted only to a very limited
extent.
With an assumed shifting time of 24 h, industrial DR provides
medium term storage. According to the calculated load reduction
potentials, between 74 GWh and 83 GWh can be stored within the
applied maximum intervention time of 4 h. Given that industrial DR
includes load shedding, only between and 29 GWh and 39 GWh of
the stored demand has to be recovered within the following 24 h.
The energy storage capacity that can be accessed by advancing
industrial loads varies between 3 GWh and 6 GWh.
In order to assess the possible contribution of DR to balancing of
intermittent power generation, the approximated storage capacity
is related to the installed capacity of renewable energies and
alternative storage options. The average load reduction potential of
93 GW is more than twice the pumped hydro storage discharging
power of 45 GW currently available in Europe [81]. The energy that

140%
120%
100%
80%
60%
40%
20%
0%

Share of minimum load increase potenƟal in peak load

Share of maximum load increase potenƟal in peak load
Share of average load increase potenƟal in peak load

Fig. 7. Share of minimum, maximum and average load increase potential in yearly peak load, subdivided by country.
 H.C. Gils / Energy 67 (2014) 1e18 13

Fig. 8. Daily load reduction average during one year for five representative technologies.

can be stored in load flexibility is however smaller than the pum-
ped hydro reservoir capacity of around 2500 GWh [82]. The average
load increase of 247 GW is much higher than the installed capac-
ities of wind and PV power, which by the end of 2011 added up to
roughly 96 GW and 51 GW, respectively. Given the ongoing
expansion of renewable power generation in Europe, the ratio of
load increase potential to intermittent capacity is expected to
decrease in the future.
In the evaluation of the findings it has to be considered that this
assessment is limited to theoretical potentials. Restrictions in DR
use resulting from the manifold technical, economic, legal and so-
cietal barriers have been disregarded completely. Additionally, the
stated potentials include also those consumers that are already
participating in DR programs, which might further reduce the
accessible DR resources. On the other hand, the usage of onsite
power generation could allow for additional grid load reductions,
which are not taken into account here. The same applies for further
consumers with demand flexibility not included in this assessment,
as well as increased industrial demand flexibility provided by the
installation of physical storage for intermediate products.
The results of this assessment rely on various assumptions and
simplifications, affecting both the average potentials and their
temporal availability throughout the year. Some of them might
cause an over- or underestimation of the overall potentials, others a
wrong distribution to or within the countries in the investigation
area. In the absence of detailed statistics, country and consumer
specific peculiarities are considered only to a minor degree.
Calculated annual energy demands of flexible consumers rely on a
number of assumptions with significant impact on the results. This
is particularly important in residential and commercial sector. The
use of European averages in the calculation of demand shares of DR
appliances in the commercial sector might relocate a certain share
of the potential from one country to another. The same goes for the
global assumptions made for the energy demands and usage
pattern of the residential washing and cooling appliances, which
were applied to all countries in the investigation area. It has not
been considered that the efficiency standard of devices varies be-
tween countries. Industrial potentials are by large extend based on
detailed production statistics. However, the assumed specific en-
ergy demands, utilization levels and minimum process loads might

Fig. 9. Hourly load reduction average during one spring week for five representative technologies.
14 H.C. Gils / Energy 67 (2014) 1e18

Fig. 10. Daily load reduction average during one year for five selected countries relative to the annual maximum.

Fig. 11. Daily load increase average during one year for five selected countries relative to the annual maximum.

not apply to the same extend to all facilities. This is also the case for
the utilization of cooling and ventilation equipment in industrial
and commercial sector.
Another important assumption is related to the load profiles.
Given that no comprehensive database of metered load was
available, exemplary load profiles found in literature were used.
Consequently, neither country-specific household activity and
appliance usage patterns, nor facility-specific industrial production
cycles are reflected in the resulting hourly potentials. A more
detailed analysis of national or regional DR potentials would have

Fig. 12. Density of the load reduction potential in kW/km2.

 H.C. Gils / Energy 67 (2014) 1e18
Fig. 13. Density of the load increase potential in kW/km2.
to rely on a broader database of metered load profiles, as well as
technology and consumption characteristics.
The geographic allocation of flexible consumers provides a first
approximation of the regional concentration of DR potentials. It is
detailed and robust for industry, since single production sites have
been identified. Also the usage of population density data for res-
idential DR loads provides a realistic allocation. In contrast to that,
the data set used for the commercial consumers represents a rough
simplification of the real demand distribution.
The primary focus of this work is to gain an overview of the
electricity consumers that might be used for demand response
and to provide a first estimate of its loads in Europe. Due to the
large number of processes, appliances and countries considered,
and the lack of country-specific consumption data, it does not
reach the degree of detail of other, more focused studies. Even
though they rely on powerful assumptions and approximations,
the results of this assessment offer an indication where high
amounts of sheddable and shiftable loads can be accessed. Ac-
cording to the results presented, it appears attractive to pursue
DR programmes in all consumer sectors. Based on this analysis,
countries and regions with particularly high DR potential can be
identified. With substantial flexible loads identified, the results
suggest that DR can in principle contribute to the stability of
energy systems with high shares of fluctuating renewable
energies.
5. Conclusion
In this paper, an assessment of the theoretical DR potential in
Europe is presented. It includes 30 different electricity consumers
across all demand sectors, which can shift or shed their load for at
least 1 h. The quantification of flexible loads in industry also
comprises the Mediterranean countries of North Africa. Special
attention is given to the evaluation of temporal availability and
geographic allocation of qualified consumers. Load profiles of the
corresponding processes and appliances are considered in order to
analyse the temporal variability of the overall DR potential. Flexible
loads in the energy-intensive industries are allocated explicitly to
the existing production sites, whereas a simplified approach is
applied in the geographic distribution of those in residential and
commercial sector. The developed methodological framework can
15
easily be adapted to other world regions or more detailed input
data.
The aggregated theoretical DR potential for all analysed coun-
tries and consumers amount to hourly averages of 93 GW for load
reduction and 247 GW for load increase are determined. Here, load
increase is equivalent to advancing the demand to an earlier time,
whereas load reduction can be either delaying or shedding. Flexible
loads are available in all demand sectors. Load increase can mainly
be realized in private households, whereas the load reduction is
distributed with similar shares to the three demand sectors. The
considered processes and appliances contribute to very different
extents to the overall potential. The average load reduction is
dominated by the steel, cement, pulp and paper industry, as well as
cooling, air conditioning and washing equipment in the commer-
cial and residential sector. Additional loads that can be activated for
DR are almost completely provided by electric space heating,
storage water boilers and washing equipment.
Due to the changing load of the flexible consumers, the poten-
tials show extensive variations during the course of the year. With
the assumed demand profiles, the reducible load varies from 61 GW
to 172 GW, the increasable load from 68 GW to 499 GW. It is shown
that the consideration of the temporal availability of flexible loads
is particularly important in the residential and commercial sector,
where in some hours of the year the summarized reduction po-
tential drops to around 20% of their respective annual maximum
value. Considering residential load increase, the minimum is even
lower and accounts for only 10% of the maximum potential. The
annual curve of available DR potential is flatter in countries with
high industrial shares. In contrast, variations are particularly pro-
nounced in countries with great amounts of electric heating and air
conditioning.
Comparing European countries, the identified potentials for DR
show clear differences in quality and quantity. The share each
sector has in the national DR potential depends on the production
capacity of the energy-intensive industries on the one hand, and
the residential equipment rates of electric heating, air conditioning
and washing appliances on the other. An indicator for the relative
share of industrial and commercial sector potentials is the overall
flexible load per inhabitant. It is particularly high in Norway,
Finland, Sweden and Luxembourg, and low in the economically less
developed countries of eastern and south-eastern Europe. A first
16 H.C. Gils / Energy 67 (2014) 1e18

Fig. 14. Average density of the load reduction potential of each Nuts-3 region in kW/km2.

appraisement of the possible contribution of DR to system stability
and balancing can be based on the ratio of average load reduction
potential to annual peak load. It ranges from 7% to 26%, with highest
values in Luxemburg, Greece, Romania and Spain.
This work provides a first assessment of the theoretical DR po-
tential in Europe. It is based on an extensive analysis of techno-
logical characteristics and load profiles of flexible consumers. In the
absence of more detailed load data, the results presented rely on a
number of estimates, which especially affect the contribution of
countries to the overall potential and the temporal availability of
flexible loads. Future research will have to draw upon a more
detailed and comprehensive database of country and technology-
specific parameters and load profiles. Despite the limitations in
data availability, the values obtained allow for an approximate
assessment of the possible role of DR in the European energy sys-
tem. Making use of the calculated load profiles, the hourly use of DR
can be further assessed in comprehensive energy system models.
There, the participation of residential and commercial consumers
in DR programs may be considered, as well as the costs caused by
the shedding and shifting of loads. A subsequent analysis of DR can
include a comparison with alternative balancing options, such as
storages, flexible power plants, electric vehicles and transmission.
The geographic allocation of the potentials presented allows for the
examination of geographical entities of different sizes in such
models.
Acknowledgements
The research for this work has been performed in the framework
of the project “Opportunities and constraints for load balancing by
energy storage, shiftable loads and electricity-driven Combined
Heat and Power in energy systems with high renewable energy

Fig. 15. Average per capita load reduction potential of each Nuts-3 region in kW.
 H.C. Gils / Energy 67 (2014) 1e18 17

shares“, financed by the German Federal Ministry of Economics and [11] Benefits of demand response in electricity markets and recommendations for
achieving them. Washington DC: United States Department of Energy; 2006.
Technology. The author thanks Yvonne Scholz, Marlene O’Sullivan
[12] Ancillotti E, Bruno R, Conti M. The role of communication systems in smart
and Michael Nast for their helpful comments on previous drafts of grids: architectures, technical solutions and research challenges. Comput
this paper. Commun 2013;36:1665e97.
[13] Kostková K, Omelina L, Kycina P, Jamrich P. An introduction to load man-
agement. Electr Power Syst Res 2013;95:184e91.
Nomenclature [14] Aghaei J, Alizadeh M-I. Demand response in smart electricity grids equipped with
Ci annual production capacity of process i, t renewable energy sources: a review. Renew Sustain Energy Rev 2013;18:64e72.
[15] Albadi MH, El-Saadany EF. A summary of demand response in electricity
dcycle,i duration of one run of appliance type i, h
markets. Electr Power Syst Res 2008;78:1989e96.
fDR maximum number of DR load interventions per year, 1/a [16] Cappers P, MacDonald J, Goldman C, Ma O. An assessment of market and
nHDD,d mumber of heating degree days on day d, e policy barriers for demand response providing ancillary services in U.S.
electricity markets. Energy Policy 2013;62:1031e9.
nFLH,i annual full load hours of appliance i, h
[17] Strbac G. Demand side management: benefits and challenges. Energy Policy
Ncycle,i annual number of runs per unit of appliance i, e 2008;36:4419e26.
NHH number of households, e [18] Torriti J, Hassan MG, Leach M. Demand response experience in Europe: pol-
Nhours total number of hours in one year, e icies, programmes and implementation. Energy 2010;35:1575e83.
[19] Hamidi V, Li F, Robinson F. Demand response in the UK’s domestic sector.
Pincrease,i potential load increase of process/appliance i, MW Electr Power Syst Res 2009;79:1722e6.
Preduction,i potential load reduction of process/appliance i, MW [20] Soares A, Gomes Á, Antunes CH. Categorization of residential electricity
Pinstalled,i installed capacity of process/appliance i, MW consumption as a basis for the assessment of the impacts of demand response
actions. Renew Sustain Energy Rev 2014;30:490e503.
Punit
installed,i Installed capacity per unit of appliance i, MW [21] UCTE system adequacy forecast 2009e2020. Brussels: Union for the Coordi-
Pcycle,i average unit load during one run of appliance i, kW nation of Transmission of Electricity; 2009.
Qd demand on day d, MWh [22] Faruqui A, Harris D, Hledik R. Unlocking the V53 billion savings from smart
meters in the EU: how increasing the adoption of dynamic tariffs could make
Qyear annual demand, MWh or break the EU’s smart grid investment. Energy Policy 2010;38:6222e31.
ri equipment rate with household appliance i, e [23] Schleich J, Klobasa M, Gölz S, Brunner M. Effects of feedback on residential
sdemand,i share of consumer class i in the annual tertiary sector electricity demand: findings from a field trial in Austria. Energy Policy
2013;61:1097e106.
demand, %
[24] Torriti J. Price-based demand side management: assessing the impacts of
sincrease,i share in unused capacity of process/appliance i that can time-of-use tariffs on residential electricity demand and peak shifting in
be activated, % Northern Italy. Energy 2012;44:576e83.
[25] Grünewald P, Torriti J. Demand response from the non-domestic sector: early
sload,i hour share in the annual electricity demand, %
UK experiences and future opportunities. Energy Policy 2013;61:423e9.
sminimum,i minimum load share relative to installed capacity, % [26] López-Rodríguez MA, Santiago I, Trillo-Montero D, Torriti J, Moreno-Munoz A.
sreduction,i share of the current load that can be reduced, % Analysis and modeling of active occupancy of the residential sector in Spain: an
srevision,i share in total hours of the year the production is halted for indicator of residential electricity consumption. Energy Policy 2013;62:742e51.
[27] Torriti J. Demand side management for the European Supergrid: occupancy
revision, % variances of European single-person households. Energy Policy 2012;44:199e
suse,i capacity utilization as share in maximum use (full load in 206.
all hours except revision outage), % [28] Demand response as a resource for the adequacy and operational reliability of
the power systems. Brussels: European Transmission System Operators; 2007.
t time, h [29] Grein A, Pehnt M. Load management for refrigeration systems: potentials and
tinterfere maximum duration of the DR load interventions, h barriers. Energy Policy 2011;39:5598e608.
tshift maximum time until shifted load has to be balanced, h [30] Shipman R, Gillott M, Naghiyev E. SWITCH: case studies in the demand side
management of washing appliances. Energy Procedia 2013;42:153e62.
Wi annual electricity demand of process/appliance i, MWh [31] Stadler I. Power grid balancing of energy systems with high renewable energy
spec
Wi specific electricity demand per output unit of process i, penetration by demand response. Util Policy 2008;16:90e8.
kWh/t [32] von Scheven A, Prelle M. Lastmanagementpotenziale in der stromintensiven
Industrie zur Maximierung des Anteils regenerativer Energien im bezogenen
Wiunit annual electricity demand per unit of appliance i, kWh
Strommix. In: Proceedings of the VDE annual conference, Stuttgart; 2012.
Wtertiary annual tertiary sector electricity demand, MWh [33] Popp D. Analysis of industrial demand response options in the German power
w temperature,  C market. In: Proceedings of the 8th international Energy economy conference,
Vienna; 2013.
[34] Database of the German Weather Service (DWD).www.dwd.de; 2010 (access
References due to payment).
[35] Prior D. Nachbildung der Energiebedarfsstruktur der privaten Haushalte e
[1] Federal Energy Regulation Commission. 2010 assessment of demand response Werkzeug zur Bewertung von Energieeinsparmaßnahmen. VDI-Verl; 1997.
and advanced metering. Washington DC: United States Department of Energy; [36] Quinkertz R. Optimierung der Energienutzung bei der Aluminiumherstellung.
2011. RWTH Aachen; 2002.
[2] Gellings CW. The concept of demand side management for electric utilities. [37] Gentilhomme P. World mining and metals yearbook e zinc. 2007 Edition.
Proc IEEE 1985;73(10):1468e70. Paris: Société de l’Industrie Minérale und BRGM Service des Ressources
[3] Bradley P, Leach M, Torriti J. A review of the costs and benefits of demand Minérales; 2007.
response for electricity in the UK. Energy Policy 2013;52:312e27. [38] Reference document on best available techniques in the non ferrous metals
[4] Barton J, Huang S, Infield D, Leach M, Ogunkunle D, Torriti J, et al. The evo- industries, integrated pollution prevention and control (IPPC). Brussels: Eu-
lution of electricity demand and the role for demand side participation, in ropean Commission; 2001.
buildings and transport. Energy Policy 2013;52:85e102. [39] Worldsteel Committee on Economic Studies. Steel statistical yearbook 2011.
[5] Critz DK, Busche S, Connors S. Power systems balancing with high penetration Brussels: World Steel Association (Worldsteel); 2011.
renewables: the potential of demand response in Hawaii. Energy Convers [40] O’Brien TF, Bommaraju TV, Hine F. Handbook of chlor-alkali technology. New
Manag 2013;76:609e19. York: Springer; 2005.
[6] Mathiesen BV, Dui N, Stadler I, Rizzo G, Guzovi Z. The interaction between [41] European chlorine industry review 2010/2011. Brussels: Euro Chlor; 2011.
intermittent renewable energy and the electricity, heating and transport [42] Schleich J, Meyer B, Lutz C, Nathani C. Endogenous technological change and
sectors. Energy 2012;48:2e4. CO2-emissions. Fraunhofer IRB Verlag; 2006.
[7] Arteconi A, Hewitt NJ, Polonara F. Domestic demand-side management [43] Blechschmidt J, editor. Altpapier: Regularien e Erfassung e Aufbereitung -
(DSM): role of heat pumps and thermal energy storage (TES) systems. Appl Maschinen und Anlagen e Umweltschutz. München: Fachbuchverlag Leipzig
Therm Eng 2013;51:155e65. im Carl-Hanser-Verlag; 2011.
[8] Beaudin M, Zareipour H, Schellenberglabe A, Rosehart W. Energy storage for [44] Blechschmidt J, editor. Taschenbuch der Papiertechnik. München: Fachbuch-
mitigating the variability of renewable electricity sources: an updated review. verlag Leipzig im Carl-Hanser-Verlag; 2010.
Energy Sustain Dev 2010;14:302e14. [45] United Nations Food and Agriculture Organization (FAO). Database on http://
[9] Kusch W, Schmidla T, Stadler I. Consequences for district heating and natural data.un.org [accessed 24.08.12].
gas grids when aiming towards 100% electricity supply with renewables. [46] Paulus M, Borggrefe F. The potential of demand-side management in energy-
Energy 2012;48:153e9. intensive industries for electricity markets in Germany. Appl Energy 2011;88:
[10] Kreith F. Energy management and conservation handbook. CRC Press; 2007. 432e41.
18 H.C. Gils / Energy 67 (2014) 1e18
[47] Verein Deutscher Zementwerke. Umweltdaten der deutschen Zementindus-
trie 2010; 2011. Düsseldorf.
[48] Cameron W. In: Kirk-Othmer, editor. Encyclopedia of chemical technology.
Hoboken: John Wiley and Sons Inc.; 2010.
[49] Gutschi C, Stigler H. Potentiale und Hemmnisse für power demand side
management in Österreich. In: Proceedings of the 10th Symposium Energy
Innovation, Graz, Austria; 2008.
[50] Häring H-W, editor. Industrial gases processing. Weinheim: Wiley-VCH; 2008.
[51] Eurostat European Statistics Database. Section nrg_10: energy statistics e
supply, transformation, consumption, dataset nrg_105a; as of November 16,
2012.
[52] International Energy Agency. Energy balances of non-OECD countries, edition
2009.
[53] International Energy Agency. Energy balances of OECD Countries, edition
2009.
[54] Baumert K, Selman M. Heating and cooling degree days; 2003.
[55] Eurostat European Statistics Database. Energy statistics e heating degree-days
(nrg_esdgr), heating degree-days by NUTS 2 regions e monthly data (nrg_
esdgr_m); as of February 4, 2011.
[56] Steimle F, Kruse H, E W, Jahn K. Energiebedarf für die technische Erzeugung
von Kälte. Hannover: Statusbericht des Deutschen Kälte- und Klimatechni-
schen Vereins; 2002.
[57] Klobasa M. Dynamische Simulation eines Lastmanagements und Integration von
Windenergie in ein Elektrizitätsnetz auf Landesebene unter regelung-
stechnischen und Kostengesichtspunkten. Stuttgart: Fraunhofer RB-Verl.; 2009.
[58] Radgen P, editor. Market study for improving energy efficiency for fans.
Stuttgart: Fraunhofer IRB Verl.; 2002.
[59] Schlomann B, Dütschke E, Gigli M, Steinbach J, Kellberger H, Geiger B, et al.
Energieverbrauch des Sektors Gewerbe, Handel, Dienstleistungen (GHD) in
Deutschland für die Jahre 2007 bis 2010. Karlsruhe, München, Nürnberg:
Bericht an das Bundesministerium für Wirtschaft und Technologie; 2011.
[60] Lange M, Focken U, Bümmerstede J, Klobasa M. Kurz- bis mittelfristige
Marktpotentiale für Demand Response-Anwendungen im gewerblichen Sek-
tor. In: Proceedings of the Vernetzungstagung, Berlin; 2010.
[61] Statistisches Jahrbuch. Red.-Schluss: 1. August 2011. Statistisches Bundesamt;
2011.
[62] Haberkern B, Maier W, Schneider U. Steigerung der Energieeffizienz auf
kommunalen Kläranlagen. Dessau-Roßlau: German Federal Environment
Agency (UBA); 2008. Report to the.
[63] Lanz M, Fricke B, Anthrakidis A, Genter M, Hoffschmidt B, Faber C, et al. CO2-
Emissionsminderung durch Ausbau, informationstechnische Vernetzung und
Netzoptimierung von Anlagen dezentraler, fluktuierender und erneuerbarer
Energienutzung in Deutschland. Dessau-Roßlau: German Federal Environ-
ment Agency (UBA); 2011. Report to the.
[64] Dena. dena-Netzstudie II. Integration erneuerbarer Energien in die deutsche
Stromversorgung im Zeitraum 2015e2020 mit Ausblick 2025. Deutsche
Energieagentur (dena); 2010.
[65] Frey G, Schulz W, Horst J, Leprich U. Energieeffizienzpotenziale durch Ersatz
von elektrischem Strom im Raumwärmbereich. Saarbrücken, Bremen: Institut
für Zukunftsenergiesysteme and Bremer Energie Institut; 2007.
[66] Hofer P. Niveau und Entwicklung des Elektrizitätsverbrauches Ohm’scher
Widerstandsheizungen in den Privaten Haushalten. Basel: Swiss Federal Of-
fice of Energy (BFE), Prognos AG; 2007. Report to the.
[67] Letschert VE, McNeil MA. Material world: forecasting household appliance
ownership in a growing global economy. In: Proceedings of the ECEEE 2009
summer study, La Colle Sur Lop; 2009. pp. 1881e7.
[68] Niro G, Salles D, Alcántara MVP, da Silva LCP. Large-scale control of domestic
refrigerators for demand peak reduction in distribution systems. Electr Power
Syst Res 2013;100:34e42.
[69] Finn P, O’Connell M, Fitzpatrick C. Demand side management of a domestic
dishwasher: wind energy gains, financial savings and peak-time load reduc-
tion. Appl Energy 2013;101:678e85.
[70] Schlomann B, Gruber E, Eichhammer W, Diekmann J, Ziesing HJ, Herzog T,
et al. Energieverbrauch der privaten Haushalte und des Sektors Gewerbe,
Handel, Dienstleistungen. Karlsruhe: Fraunhofer ISI; 2004.
[71] Enerdata Research Service. Odyssee database on energy efficiency data and
indicators; 2012.
[72] Bertoldi P, Atanasiu B. Electricity consumption and efficiency trends in the
European Union e status report 2009. Ispra: Joint Research Center, Institute
for Energy; 2009.
[73] Seebach D, Timpe C, Bauknecht D. Costs and benefits of smart appliances in
europe, report prepared as part of the EIE project ’Smart domestic appliances
in Sustainable Energy systems (Smart-A). Freiburg: Öko-Institut; 2009.
[74] Wohlauf G, Thomas S, Irrek W, Hohmeyer O. Ersatz von Elektro-
Speicherheizungen durch effiziente Brennwerttechnik. Wuppertal: Wupper-
tal Institut für Klima, Umwelt, Energie GmbH; 2005.
[75] Stadler I. Demand response: nichtelektrische Speicher für Elek-
trizitätsversorgungssysteme mit hohem Anteil erneuerbarer Energien; 2006.
dissertation.de.
[76] Stamminger R, Broil G, Pakula C, Jungbecker H, Braun M, Rüdenauer I, et al.
Synergy potential of smart appliances. University of Bonn; 2008.
[77] Minerals yearbookVolumes report international, vol. 3. United States
Geological Survey; 2010.
[78] European Environmental Agency. CORINE land cover raster data, version 13;
2010.
[79] Gallego FJ. A population density grid of the European Union. Popul Environ
2009;31:460e73.
[80] International Earth Science Information Network (CIESIN) CU Center for,
Agricultura Tropical (CIAT) CI deGridded population of the world, Version 3
(GPWv3); 2005.
[81] Eurostat European Statistics Database. Section nrg_11: energy statistics e
infractructure, dataset nrg_113a; as of November 26, 2013.
[82] Pirker O, Argyrakis I, Babkin V, Chudy M, Crosnier G, Dahlback N, et al. Hydro
power in europe: powering renewables. Eurelectric; 2011.