Professional Documents
Culture Documents
Assessment of The Theoretical Demand Response Potential in Europe
Assessment of The Theoretical Demand Response Potential in Europe
Assessment of The Theoretical Demand Response Potential in Europe
Energy
journal homepage: www.elsevier.com/locate/energy
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 thermal storages in combined heat and power or heat pump sys-
tems [7e9]. Modifications in the demand pattern are typically
DR (Demand Response) actions are defined as “changes in realized by direct or indirect load management programmes [10].
electric use by demand-side resources from their normal con- Existing DR measures include time-based rates on one hand, and
sumption patterns in response to changes in the price of electricity, incentive based programmes on the other [11]. Technical require-
or to incentive payments designed to induce lower electricity use at ment for the participation in DR programs is the availability of an
times of high wholesale market prices or when system reliability is information and communication infrastructure allowing for the
jeopardized” [1]. In contrast to demand side management, which transmission of and reaction to load, price and control signals.
also comprises energy efficiency measures and permanent and/or Communication channels include radio and telecommunication, as
regular utility-driven changes in the demand pattern, DR is focused well as power lines [12,13]. Markets for flexible loads range from
on load flexibility and short term costumer action [2,14]. In elec- on-site peak load reduction and increased internal PV consumption
tricity systems primarily relying on fossil and nuclear fuels, in- to the participation in energy trade, the provision of operating
terventions in customer load can increase the profitability of power energy, as well as the clearance of imbalances in the TSO (trans-
plants and help to avoid investments in additional generation or mission system operator) area and the management of supply
grid capacities [3,17]. Most demand side measures are thus shortfalls [14,15]. Due to various technical, economic, legal and
designed to either enable load curtailments in times of peak de- societal barriers, the use of DR is limited so far [16,17]. Existing
mand or to shift loads to times of low demand [13]. In supply programmes are mostly focusing on large industrial consumers,
systems with high capacities of intermittent solar PV(photovoltaic) however, residential and commercial demand is increasingly also
and wind power, the flexibility provided by DR can create addi- taken into consideration [18].
tional benefits [4e6]. By adjusting the demand to the present Demand response resources have been identified in a broad
availability of fluctuating resources, curtailments can be reduced range of processes and devices [19,20,25,27,31,57]. In Germany and
and the overall RE (renewable energy) share can be increased. With Austria, their average shiftable and sheddable loads add up to
shifting times ranging from some minutes to a few days, DR com- several GW [29,31,57]. The ENTSO-E (European Network of Trans-
petes with alternative balancing options, such as batteries and mission System Operators for Electricity) has quantified a load
pumped storage power stations, as well as grid extension, but also 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-
* Tel.: þ49 711 6862 477; fax: þ49 711 6862 747. ieved by smart meter installation and dynamic pricing on a Euro-
E-mail address: hans-christian.gils@dlr.de. 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- 2.1. Identification of shiftable loads and required parameters
bution of residential appliances has been investigated in field trials
realized in Austria and Northern Italy [23,24]. In this study, a total of 30 different processes and appliances are
A consumer and country-specific analysis of the flexible loads on taken into consideration. Shiftable loads typically feature one of the
the European continent is missing so far. In this paper, the theo- following characteristics: heat or cold storage (e.g. space heating,
retical DR potential in Europe and North Africa is assessed.1 Its aims refrigerators), demand flexibility (e.g. washing, ventilation) or
are the characterization of electricity consumers that are able to physical storages (e.g. cement industry, fresh water supply). In-
shift or shed their load for a given period of time and the provision of dustrial load shifting may be limited by technical constraints,
a first estimate of their loads in Europe. Three different DR actions process requirements and availability of unutilized plant or ma-
are taken into account: load shedding, load shifting to an earlier time chine capacity. For processes with very high utilization rates e as
and load shifting to a later time. Network load reductions achieved they are found in energy-intensive industries e only load shedding
by the usage of costumer-owned onsite generation are not included without previous or subsequent balancing can be implemented. In
in the analysis. DR potentials are determined across all demand residential and commercial sector, typically both load shifting and
sectors: industry, as well as the tertiary and residential sector. shedding can be realized. Due to higher costs and losses of comfort,
When assessing the potential future contribution of DR to the this study evaluates load shedding only for energy-intensive in-
system integration of RE, the temporal availability of flexible loads dustrial processes. Table 1 provides an overview of processes and
is of particular importance [25]. The demand response behaviour of appliances included. The analysis is limited to those loads that can
non-residential consumers is directly correlated to industrial pro- be shifted or shedded for at least 1 h. Detailed descriptions of their
duction activity and business hours. In the residential sector, time- technical properties and DR behaviour can be gathered from the
related electricity demands can be derived from an evaluation of references cited in Table 1.
household activity level and occupancy variance [26,27]. In the For each country, DR consumer and hour of the year, potential
work presented here, exemplary load profiles of all relevant con- load increase Pincrease and load reduction Preduction are assessed.
sumers are either estimated or extracted from metered data Load reduction can be realized either by shedding or delaying de-
available in literature. Based on these profiles, potentials for load mand, whereas load increase is equivalent to advancing the oper-
reduction and increase are calculated for each hour of the year. In ation of processes or devices. Loads can thus be shifted to both an
the context of the balancing of RE fluctuations, also the duration of earlier or later point in time. In case of load shifting, every load
load interventions, the shifting time and the frequency of DR ac- increase is followed by a decrease due and vice versa. Flexible loads
tions are of special importance. These parameters have decisive are calculated from characteristic load profiles and annual elec-
impact on the quality of the corresponding DR potential. In order to tricity demands. Furthermore, shares of used and free capacity that
facilitate follow-on studies of the interaction of DR measures with can be activated for DR are assessed. Annual demands are obtained
the electricity grid utilization, also the geographical distribution of from statistics or estimated based on industrial production capac-
flexible consumers is evaluated. Given that the application of DR is ities and equipment rates of domestic appliances. Given that for the
subject to constraints, different kinds of potentials need to be North African countries no tertiary sector electricity demands and
treated separately. It can be distinguished between the theoretical, appliance equipment rates are available, only potentials in energy
technical, economic and practical potential [28,29]. Whereas the intensive industries are considered there.
theoretical potential comprises all facilities and devices of the Power demand flexibility can be compared to a functional en-
consumers suitable for DR, the technical potential includes only ergy storage with limited storage period. Its charging capacity is
those that can be controlled by the existing information and determined by the flexible load, its reservoir capacity by the
communication infrastructure. A subset of the technical potential is maximum duration of DR interventions tinterfere, and its maximum
the economic potential of all DR consumers that can be operated in storage period by the shifting time tshift. The shifting time defines
a cost-efficient way. Another independent subset of the technical the maximum duration until load that has been advanced or
potential arises from the acceptance of load interventions. The delayed needs to be balanced again, whereas the intervention time
effectively usable, practical potential consists of the intersection of reflects a limit in duration of changes in the normal demand
economic potential and accepted use. This work is restricted to the pattern. Taking into account an annual limit in number of DR in-
assessment of the theoretical DR potential. Limitations for technical terventions fDR, the storable energy per year can be calculated.
reasons not related to industrial production processes, costs or Parameters limiting DR are typically depending on process cycles,
refusal to participate will at this point be neglected. physical storage capacities for intermediate products or the ther-
mal capacity of heated/cooled goods or rooms. The assumed values
2. Data and methodology for interference and shift times, as well as frequency of DR events
are summarized in Table 1.
The analysis is performed in four steps. First, processes and
appliances suitable for DR are identified (Section 2.1). Then, their 2.2. Demand profiles of shiftable loads
characteristic load profiles are assessed (2.2). In the third step,
annual electricity demands and installed capacities in the year 2010 In order to analyse the variability of the DR potential during the
are quantified, and a flexible load share for each consumer is year, exemplary load profiles are taken into account. As no own
evaluated (2.3e2.5). Finally, the geographical distribution of DR measurements have been performed, metered data and informa-
potentials is investigated (2.6). 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
1
Countries that are included: Albania, Algeria, Austria, Belarus, Belgium, Bosnia on energy usage, the load profile is assumed to follow characteristic
and Herzegovina, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Egypt, periodic seasonal, weekly and daily profiles. For technologies
Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Libya, providing heat or cold, hourly demands are further correlated to
Liechtenstein, Lithuania, Luxembourg, Macedonia, Malta, Moldova, Montenegro,
Morocco, Netherlands, Norway, Poland, Portugal, Romania, Serbia and Kosovo,
outside temperature. Whether the electricity demand is assumed
Slovakia, Slovenia, Spain, Sweden, Switzerland, Tunisia, Turkey, United Kingdom, to depend on the time or temperature is indicated in Table 1 for
Ukraine. 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.
e h h 1/a e e
2.2.1. Constant and periodic seasonal, weekly and daily profiles days (HDD/CDD) are calculated for each country and, where avail-
Energy-intensive production processes are typically running at able, Nuts-2 statistical region.
very high capacity utilization levels [46]. For this reason, a constant Daily shares Qd in annual space heating demand Qyear are
load is applied during all hours of the year. Only exception is the calculated with an extended degree day method, which considers
cement industry where utilization ranges between 40% and 100%. the HDD number nHDD,d not only of the current, but also the six
In addition to winter times e when construction activities are preceding days (see Equation (1)). In doing so, the HDDs in the
typically reduced e production is also lowered in the daytime on Nuts-2 regions of each country are weighted according to its pop-
workdays [32,46]. It is assumed that utilization in winter is by 20% ulation share. The hourly demand during each day is estimated
lower than in summer, and in the daytime on workdays at all using temperature-dependent average profiles, which were
seasons reduced to two thirds of its night load. For industrial derived from the network load profile of a German district heating
ventilation energy demand, a weekend decline of 40% (Saturday) supplier of residential and commercial consumers.
and 50% (Sunday) is assumed; commercial ventilation is further- P 1
more reduced by 50% at night-time. The electricity demand of 2a nHDD;da
a
cooling appliances in private homes, retailing, hotels and restau- Qd ¼ P 1 P365 $Qyear a ¼ 0.6 ðMWhÞ (1)
rants is estimated to be by 10% lower in winter times than in 2a $ d ¼ 1 nHDD;d
a
summer; it additionally declines by 20% at night, given that the
frequency of user interventions tends to go down. Based on With a difference of only 20% between coldest and warmest day
metered data presented in Ref. [33], cold storages and industrial of the year, daily hot water demand is assumed to be almost con-
cooling are assumed to have lower demands on Saturdays (5%), stant. This assumption is derived from the analysis of the same
Sundays (10%) and during peak network load hours in the district heating time series. The hourly demand within each day
morning (50%). Pumps in the fresh water supply are also typically follows the measured load profile presented in Ref. [35].
running during off-peak hours at night; here it is assumed that load Daily shares in annual air conditioning demand are also calcu-
is reduced by two thirds in the daytime. The operation of washing lated with Equation (1). For cooling demand, however, only the
machines, tumble dryers and dish washers is mainly driven by the CDD of the current and previous day are taken into consideration.
daily routine of its users. Ref [35] provides measured hourly usage Within each day, air conditioning power demand is assumed to
profiles for different weekdays and seasons, which are used here. peak in afternoon hours, in accordance with the representative
profile shown in Ref. [76].
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 (2) 2.3.2. Industrial cross-sectional technologies
Flexible loads in the cross-sectional technologies cooling and
Wi ventilation are evaluated based on their annual electricity demand
Pinstalled;i ¼ ðMWÞ (3) Wi. These demands are estimated using data from Refs. [56,58,71].
Nhours $suse;i $ 1 srevision;i
Dividing Wi by the number of full load hours nFLH, the installed
The potential load reduction Preduction,i (t) in each hour is given capacity is obtained (see Equation (6)). In contrast to energy-
by the difference between current load and minimum load of the intensive processes, no revision outage is considered.
process (see Equation (4)). Its value changes during the year ac-
cording to the load profile sload,i (t). The minimum process load is Wi
Pinstalled;i ¼ ðMWÞ (6)
defined relative to the installed electrical capacity and given by nFLH;i
parameter sminimum,i. The potential load increase Pincrease,i (t) is
In the assessment of potential load reduction and increase, fixed
calculated from the difference between maximum load and current
shares in current load sreduction,i and unused capacity sincrease,i
load, which is at least temporarily greater than zero for all pro-
available for DR are assumed. Taking into account the load in the
cesses operated at less than 100% utilization. This difference is
current and subsequent hours, they allow for the calculation of
multiplied with parameter sincrease,i, reflecting the free production
shiftable loads according to Equations (7) and (8). In case of a load
capacity share available for DR (see Equation (5)). Table 2 sum-
increase, all demand of the following tshift hours can be advanced.
marizes the assumed parameter values.
Its upper limit is set by the installed capacity. Estimated energy
demands, utilization levels and DR shares of the industrial cross-
Preduction;i ðtÞ ¼ sload;i ðtÞ$Wi Pinstalled;i $sminimum;i ðMWÞ (4)
|fflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} sectional technologies are summarized in Table 3.
Load in hour t Minimum Load
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 % % % %
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 With the annual electricity demands and full load hours sum-
marized in Table 6, the installed capacity is calculated according to
Tertiary sector DR potentials are available in the supply of cold, Equation (6). No procedural limitations of load shifting are
heat, water and ventilation, as well as in waste water treatment. considered in tertiary sector. Load shares available for DR are thus
Flexible loads in these applications are calculated based on their set to 100% of current load and unused capacity, respectively. Based
annual energy consumptions. In the absence of country-specific on annual demand, installed capacity and hourly load profiles,
data, they are approximated by multiplying the tertiary sector de- possible load reduction and increase are calculated according to
mand [51e53] with average demand shares of the relevant uses. Equations (7) and (8), respectively.
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, 2.5. Flexible loads in the residential sector
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 Residential load shifting is evaluated for heating, cooling, air
conditioning, which are assumed to depend on outside tempera- conditioning and washing equipment. The latter comprises
ture, these shares are applied to all European countries. Pursuant to washing machines, tumble dryers and dish washers. In contrast to
values estimated for Germany in Ref. [56], the electricity demand of the other sectors, residential DR loads are quantified in a bottom-up
cooling appliances is subdivided to food retailing (75%), cold stor- approach. Household numbers and country-specific equipment
ages (10%) and hotels/restaurants (15%). rates ri of relevant devices are obtained from Refs. [64e67,71,73], or
approximated using available data (see Table 7). Multiplying the
Wi ¼ Wtertiary $sdemand;i ðMWhÞ (9) unit number with the specific capacity and energy consumption,
country sums are calculated for each appliance type according to
The air conditioning share in the sectoral demand is approxi- Equations (10) and (11). Based on [68,72], an annual refrigerator
mated using long-term average CDDs of each country [54]. It ranges and freezer unit demand of 350 kWh is applied.
between 0.5% in northern countries to 12% on the Mediterranean
islands (see Table 4 and Table 7). These values rely on shares unit
Pinstalled;i ¼ NHH $ri $Pinstalled;i ðMWÞ (10)
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 Wi ¼ NHH $ri $Wiunit ðMWhÞ (11)
are estimated for each country based on long-term average HDDs For washing equipment, average power demands during use, as
(see Table 7) [55]. It is assumed, that the usage is higher in colder well as frequency and duration of use are taken into account (see
climates, as to the values in Table 5. In the assessment of air Table 8). Relying on measured consumption data from Refs. [69,70],
conditioner and residential heat circulation pump full load hours, in annual demands are calculated (Equation (12)).
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 Wiunit ¼ Pcycle;i $dcycle;i $Ncycle;i ðMWhÞ (12)
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, Wiunit ¼ Pinstalled;i
unit
$nFLH
i ðMWhÞ (13)
capacity utilization reaches 1.25-times the ratio between current
In the calculation of annual space heating, hot water generation
and peak load. Resulting full load hours range between 136 in
and air conditioning electricity consumption, unit capacities are
Estonia and 991 in Cyprus, with a European average of 467 h (see
multiplied with estimated full load hours (Equation (13)). Average
Table 7). In countries without CDD, it is set to 100 h/year.
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].
Table 4
Air conditioning (AC) share in commercial electricity demand.
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
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.
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
140000
MW
120000 Load Increase Min 202297
100000
Load Increase Max
80000
60000 Load Increase Average
40000
20000
0
conditioning, as well as residential washing equipment. Fig. 8 il- reduction and load increase, respectively, in five representative
lustrates the development of the daily load reduction average European countries, all normalized to the corresponding maximum
during one year for five representative technologies. It reflects the potential. In Finland, where the industrial share in the potential is
load profiles assumed in Section 2.2. With no load changes very high and almost no air conditioning is used, daily averages of
considered, DR potentials in energy-intensive industries e here the load reduction remain almost constant during the year. In contrast,
aluminium electrolysis is selected as example e are constantly strong summer peaks can be observed in Spain and Italy. Due to a
available throughout the whole year. Also the power demand of more frequent use of washing and heating equipment in winter
food cooling appliances in all sectors shows only minor variations. times, the potential load increase is relatively higher in winter. This
In contrast, shiftable loads in the provision of air conditioning and tendency is at hand in all countries considered here; its exact shape
space heating are strongly influenced by outside temperature and depends on the installed capacity of electric space and water
have annual load curves contrasting each other. Particularly in air heating systems, as well as the course of the heating season.
conditioning demand, where short-term reactions to temperature
occur, high peaks in single days and hours can be observed. The 3.3. Geographic allocation of flexible loads
variations in the profile of the load reduction potential of com-
mercial ventilation are directly correlated to the assumed weekend DR potentials are concentrated to centres of population and
demand decline. Due to the high air conditioner share in overall energy intensive industry production. Fig. 12 and Fig. 13 show the
theoretical load reduction potential, the annual maximum is potential load reduction and increase density in each grid cell.
reached in summer times. In peak hours, the potential can be twice Major cities can be easily identified there.
as high as on average. Hours with lowest reduction potential are For further analysis of the spatial allocation of flexible loads,
found in the transition period between winter and summer, when sums over each Nuts-3 statistical region are formed, and average
both air conditioner and heat circulation use is low. Average daily values per km2 and inhabitant are calculated. The regional density
reduction potentials are found to be lower on weekends than on of the load reduction potential is displayed in Fig. 14. It reaches high
working days. This results from reduced industrial and commercial values not only in densely populated regions, but also those with a
demand. concentration of energy intensive energies. The highest values of
The flexible load is varying considerably also within each day. more than 900 kW/km2 are found in Paris, Inner London and the
This is illustrated for a selection of consumers and a representative industrial city of Ludwigshafen am Rhein in Germany. Compara-
spring week in Fig. 9. Again, the DR potential provided by industry tively low densities are present in sparsely populated areas, for
and cooling appliances is fairly constant throughout the day. The example in north-eastern Germany, Scotland or northern Finland,
same applies to residential heat circulation pumps. In contrast to Norway and Sweden.
that, the DR availability of washing equipment, air conditioners and Taking into account population density, regions with high in-
fans is heavily fluctuating. Due to the typical utilization cycles dustrial and commercial DR potentials can be identified. In Fig. 15,
driven by daylight, working hours and temperature, potentials are the per capita load reduction is shown for each region. Compara-
mostly available during daytime. This coincides with the overall tively high values are found for example in the French region of
system load pattern. Aquitaine, the Norwegian coast and Luxembourg. For most Non-EU
The identified load increase is dominated by washing and countries, no regional data was available. The figure consequently
heating equipment. Its power demands are fluctuating due to provides national average values. They are particularly low in the
different usage intensity and dependency on outside temperature. North African countries, where only potentials in the energy-
With a shifting duration of 12 h and typical night-time peaks, the intensive industries have been considered. Given that most in-
potential load increase of storage heaters is highest in evening crease potential through advancing load is provided by residential
hours. Using the demand profiles introduced in Section 2.2, the appliances, the geographic distribution is very similar to the pop-
overall load increase potential is higher in winter than in summer, ulation density. Inhabitant-specific values are almost homogenous
on weekends than on working days and at noon than during night- within all countries.
time.
The temporal availability of load flexibility arises from the 4. Discussion
composition of the overall DR potential. Differences between
countries are primarily associated to the fraction of space heating The analysis reveals a substantial theoretical DR potential in
and air conditioning. Fig. 10 and Fig. 11 include profiles for load 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
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
93 GW can be achieved. This average value is equivalent to one ventilation, heat circulation pumps, fresh water supply and
third of the 280 GW minimum load, and approximately one sev- wastewater treatment. All these appliances can be interrupted for
enth of the 620 GW peak load measured in the European part of the only one or 2 h, and the reduced load typically needs to be recov-
investigation area in 2010. The average load increase through ered within the same time span. This limits the storage period of
bringing forward electricity demand adds up to 247 GW. This very the functional storage provided by DR. Consequently, the field of
high value results from the assumption that demand of various application of these DR consumers in energy systems with high
hours can be accumulated to only 1 h. The minimum DR potentials renewable shares is restricted to the balancing of short term fluc-
available in every hour of the year amount to 61 GW of load tuations. Depending on the hour of the year, the energy that within
reduction and 68 GW of load increase. 2 h can be charged into the virtual storage by load reduction ranges
In the evaluation of the calculated theoretical potentials, it from 47 GWh to 141 GWh. The temporal variation in storage ca-
needs to be taken into account that both load and DR potential of pacity is correlated to the assumed load profiles of consumers
subsequent hours are influenced, whenever a load reduction or participating in DR. The discharging has to begin immediately af-
increase is called. All shifted demand needs to be balanced within a terwards and is also limited to a duration of 2 h. The amount of
given period of time, thus increasing or decreasing the load in one energy that can be stored by advancing the short-term DR loads is
or several following hours. Given that one and the same load cannot smaller, and varies from 4 GWh to 12 GWh.
be brought forward twice, the load increase potential of different Longer intervention times are found for washing equipment, as
hours is correlated to each other. Due to limitations in duration and well as residential and commercial electric space and water
frequency of DR interventions, this is also the case for the load heaters. With the assumed charging durations of up to 12 h for
reduction potential. electric heat production and 6 h for washing equipment, the energy
The load shifting potential is dominated by customers allowing storage capacity is comparatively high. Due to its dependency on
only short interventions and shifting times (see Table 1). They outside temperature and appliance usage pattern, it shows strong
include cooling processes in all sectors, air conditioning, 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 With an assumed shifting time of 24 h, industrial DR provides
up to 217 GWh during daytime and goes down to zero at night. The medium term storage. According to the calculated load reduction
much better storage function of washing and heating equipment potentials, between 74 GWh and 83 GWh can be stored within the
comes along with shortcomings concerning efficiency and accep- applied maximum intervention time of 4 h. Given that industrial DR
tance. Due to high exergy losses, the use of electricity for the includes load shedding, only between and 29 GWh and 39 GWh of
resistive generation of heat is unfavourable compared to alterna- the stored demand has to be recovered within the following 24 h.
tives, such as district heating or condensing boilers. Even though The energy storage capacity that can be accessed by advancing
the use of resistive electric air and water heaters is likely to be industrial loads varies between 3 GWh and 6 GWh.
reduced in the coming decades in the framework of energy effi- In order to assess the possible contribution of DR to balancing of
ciency efforts, existing units can deliver an important contribution intermittent power generation, the approximated storage capacity
to DR programs. In contrast to electric storage space and water is related to the installed capacity of renewable energies and
heaters, shifting of washing equipment entails significant con- alternative storage options. The average load reduction potential of
straints for users and might be accepted only to a very limited 93 GW is more than twice the pumped hydro storage discharging
extent. power of 45 GW currently available in Europe [81]. The energy that
140%
120%
100%
80%
60%
40%
20%
0%
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- The results of this assessment rely on various assumptions and
ped hydro reservoir capacity of around 2500 GWh [82]. The average simplifications, affecting both the average potentials and their
load increase of 247 GW is much higher than the installed capac- temporal availability throughout the year. Some of them might
ities of wind and PV power, which by the end of 2011 added up to cause an over- or underestimation of the overall potentials, others a
roughly 96 GW and 51 GW, respectively. Given the ongoing wrong distribution to or within the countries in the investigation
expansion of renewable power generation in Europe, the ratio of area. In the absence of detailed statistics, country and consumer
load increase potential to intermittent capacity is expected to specific peculiarities are considered only to a minor degree.
decrease in the future. Calculated annual energy demands of flexible consumers rely on a
In the evaluation of the findings it has to be considered that this number of assumptions with significant impact on the results. This
assessment is limited to theoretical potentials. Restrictions in DR is particularly important in residential and commercial sector. The
use resulting from the manifold technical, economic, legal and so- use of European averages in the calculation of demand shares of DR
cietal barriers have been disregarded completely. Additionally, the appliances in the commercial sector might relocate a certain share
stated potentials include also those consumers that are already of the potential from one country to another. The same goes for the
participating in DR programs, which might further reduce the global assumptions made for the energy demands and usage
accessible DR resources. On the other hand, the usage of onsite pattern of the residential washing and cooling appliances, which
power generation could allow for additional grid load reductions, were applied to all countries in the investigation area. It has not
which are not taken into account here. The same applies for further been considered that the efficiency standard of devices varies be-
consumers with demand flexibility not included in this assessment, tween countries. Industrial potentials are by large extend based on
as well as increased industrial demand flexibility provided by the detailed production statistics. However, the assumed specific en-
installation of physical storage for intermediate products. 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 available, exemplary load profiles found in literature were used.
the utilization of cooling and ventilation equipment in industrial Consequently, neither country-specific household activity and
and commercial sector. appliance usage patterns, nor facility-specific industrial production
Another important assumption is related to the load profiles. cycles are reflected in the resulting hourly potentials. A more
Given that no comprehensive database of metered load was detailed analysis of national or regional DR potentials would have
to rely on a broader database of metered load profiles, as well as easily be adapted to other world regions or more detailed input
technology and consumption characteristics. data.
The geographic allocation of flexible consumers provides a first The aggregated theoretical DR potential for all analysed coun-
approximation of the regional concentration of DR potentials. It is tries and consumers amount to hourly averages of 93 GW for load
detailed and robust for industry, since single production sites have reduction and 247 GW for load increase are determined. Here, load
been identified. Also the usage of population density data for res- increase is equivalent to advancing the demand to an earlier time,
idential DR loads provides a realistic allocation. In contrast to that, whereas load reduction can be either delaying or shedding. Flexible
the data set used for the commercial consumers represents a rough loads are available in all demand sectors. Load increase can mainly
simplification of the real demand distribution. be realized in private households, whereas the load reduction is
The primary focus of this work is to gain an overview of the distributed with similar shares to the three demand sectors. The
electricity consumers that might be used for demand response considered processes and appliances contribute to very different
and to provide a first estimate of its loads in Europe. Due to the extents to the overall potential. The average load reduction is
large number of processes, appliances and countries considered, dominated by the steel, cement, pulp and paper industry, as well as
and the lack of country-specific consumption data, it does not cooling, air conditioning and washing equipment in the commer-
reach the degree of detail of other, more focused studies. Even cial and residential sector. Additional loads that can be activated for
though they rely on powerful assumptions and approximations, DR are almost completely provided by electric space heating,
the results of this assessment offer an indication where high storage water boilers and washing equipment.
amounts of sheddable and shiftable loads can be accessed. Ac- Due to the changing load of the flexible consumers, the poten-
cording to the results presented, it appears attractive to pursue tials show extensive variations during the course of the year. With
DR programmes in all consumer sectors. Based on this analysis, the assumed demand profiles, the reducible load varies from 61 GW
countries and regions with particularly high DR potential can be to 172 GW, the increasable load from 68 GW to 499 GW. It is shown
identified. With substantial flexible loads identified, the results that the consideration of the temporal availability of flexible loads
suggest that DR can in principle contribute to the stability of is particularly important in the residential and commercial sector,
energy systems with high shares of fluctuating renewable where in some hours of the year the summarized reduction po-
energies. tential drops to around 20% of their respective annual maximum
value. Considering residential load increase, the minimum is even
5. Conclusion lower and accounts for only 10% of the maximum potential. The
annual curve of available DR potential is flatter in countries with
In this paper, an assessment of the theoretical DR potential in high industrial shares. In contrast, variations are particularly pro-
Europe is presented. It includes 30 different electricity consumers nounced in countries with great amounts of electric heating and air
across all demand sectors, which can shift or shed their load for at conditioning.
least 1 h. The quantification of flexible loads in industry also Comparing European countries, the identified potentials for DR
comprises the Mediterranean countries of North Africa. Special show clear differences in quality and quantity. The share each
attention is given to the evaluation of temporal availability and sector has in the national DR potential depends on the production
geographic allocation of qualified consumers. Load profiles of the capacity of the energy-intensive industries on the one hand, and
corresponding processes and appliances are considered in order to the residential equipment rates of electric heating, air conditioning
analyse the temporal variability of the overall DR potential. Flexible and washing appliances on the other. An indicator for the relative
loads in the energy-intensive industries are allocated explicitly to share of industrial and commercial sector potentials is the overall
the existing production sites, whereas a simplified approach is flexible load per inhabitant. It is particularly high in Norway,
applied in the geographic distribution of those in residential and Finland, Sweden and Luxembourg, and low in the economically less
commercial sector. The developed methodological framework can 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 can be further assessed in comprehensive energy system models.
and balancing can be based on the ratio of average load reduction There, the participation of residential and commercial consumers
potential to annual peak load. It ranges from 7% to 26%, with highest in DR programs may be considered, as well as the costs caused by
values in Luxemburg, Greece, Romania and Spain. the shedding and shifting of loads. A subsequent analysis of DR can
This work provides a first assessment of the theoretical DR po- include a comparison with alternative balancing options, such as
tential in Europe. It is based on an extensive analysis of techno- storages, flexible power plants, electric vehicles and transmission.
logical characteristics and load profiles of flexible consumers. In the The geographic allocation of the potentials presented allows for the
absence of more detailed load data, the results presented rely on a examination of geographical entities of different sizes in such
number of estimates, which especially affect the contribution of models.
countries to the overall potential and the temporal availability of
flexible loads. Future research will have to draw upon a more Acknowledgements
detailed and comprehensive database of country and technology-
specific parameters and load profiles. Despite the limitations in The research for this work has been performed in the framework
data availability, the values obtained allow for an approximate of the project “Opportunities and constraints for load balancing by
assessment of the possible role of DR in the European energy sys- energy storage, shiftable loads and electricity-driven Combined
tem. Making use of the calculated load profiles, the hourly use of DR 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- [65] Frey G, Schulz W, Horst J, Leprich U. Energieeffizienzpotenziale durch Ersatz
trie 2010; 2011. Düsseldorf. von elektrischem Strom im Raumwärmbereich. Saarbrücken, Bremen: Institut
[48] Cameron W. In: Kirk-Othmer, editor. Encyclopedia of chemical technology. für Zukunftsenergiesysteme and Bremer Energie Institut; 2007.
Hoboken: John Wiley and Sons Inc.; 2010. [66] Hofer P. Niveau und Entwicklung des Elektrizitätsverbrauches Ohm’scher
[49] Gutschi C, Stigler H. Potentiale und Hemmnisse für power demand side Widerstandsheizungen in den Privaten Haushalten. Basel: Swiss Federal Of-
management in Österreich. In: Proceedings of the 10th Symposium Energy fice of Energy (BFE), Prognos AG; 2007. Report to the.
Innovation, Graz, Austria; 2008. [67] Letschert VE, McNeil MA. Material world: forecasting household appliance
[50] Häring H-W, editor. Industrial gases processing. Weinheim: Wiley-VCH; 2008. ownership in a growing global economy. In: Proceedings of the ECEEE 2009
[51] Eurostat European Statistics Database. Section nrg_10: energy statistics e summer study, La Colle Sur Lop; 2009. pp. 1881e7.
supply, transformation, consumption, dataset nrg_105a; as of November 16, [68] Niro G, Salles D, Alcántara MVP, da Silva LCP. Large-scale control of domestic
2012. refrigerators for demand peak reduction in distribution systems. Electr Power
[52] International Energy Agency. Energy balances of non-OECD countries, edition Syst Res 2013;100:34e42.
2009. [69] Finn P, O’Connell M, Fitzpatrick C. Demand side management of a domestic
[53] International Energy Agency. Energy balances of OECD Countries, edition dishwasher: wind energy gains, financial savings and peak-time load reduc-
2009. tion. Appl Energy 2013;101:678e85.
[54] Baumert K, Selman M. Heating and cooling degree days; 2003. [70] Schlomann B, Gruber E, Eichhammer W, Diekmann J, Ziesing HJ, Herzog T,
[55] Eurostat European Statistics Database. Energy statistics e heating degree-days et al. Energieverbrauch der privaten Haushalte und des Sektors Gewerbe,
(nrg_esdgr), heating degree-days by NUTS 2 regions e monthly data (nrg_ Handel, Dienstleistungen. Karlsruhe: Fraunhofer ISI; 2004.
esdgr_m); as of February 4, 2011. [71] Enerdata Research Service. Odyssee database on energy efficiency data and
[56] Steimle F, Kruse H, E W, Jahn K. Energiebedarf für die technische Erzeugung indicators; 2012.
von Kälte. Hannover: Statusbericht des Deutschen Kälte- und Klimatechni- [72] Bertoldi P, Atanasiu B. Electricity consumption and efficiency trends in the
schen Vereins; 2002. European Union e status report 2009. Ispra: Joint Research Center, Institute
[57] Klobasa M. Dynamische Simulation eines Lastmanagements und Integration von for Energy; 2009.
Windenergie in ein Elektrizitätsnetz auf Landesebene unter regelung- [73] Seebach D, Timpe C, Bauknecht D. Costs and benefits of smart appliances in
stechnischen und Kostengesichtspunkten. Stuttgart: Fraunhofer RB-Verl.; 2009. europe, report prepared as part of the EIE project ’Smart domestic appliances
[58] Radgen P, editor. Market study for improving energy efficiency for fans. in Sustainable Energy systems (Smart-A). Freiburg: Öko-Institut; 2009.
Stuttgart: Fraunhofer IRB Verl.; 2002. [74] Wohlauf G, Thomas S, Irrek W, Hohmeyer O. Ersatz von Elektro-
[59] Schlomann B, Dütschke E, Gigli M, Steinbach J, Kellberger H, Geiger B, et al. Speicherheizungen durch effiziente Brennwerttechnik. Wuppertal: Wupper-
Energieverbrauch des Sektors Gewerbe, Handel, Dienstleistungen (GHD) in tal Institut für Klima, Umwelt, Energie GmbH; 2005.
Deutschland für die Jahre 2007 bis 2010. Karlsruhe, München, Nürnberg: [75] Stadler I. Demand response: nichtelektrische Speicher für Elek-
Bericht an das Bundesministerium für Wirtschaft und Technologie; 2011. trizitätsversorgungssysteme mit hohem Anteil erneuerbarer Energien; 2006.
[60] Lange M, Focken U, Bümmerstede J, Klobasa M. Kurz- bis mittelfristige dissertation.de.
Marktpotentiale für Demand Response-Anwendungen im gewerblichen Sek- [76] Stamminger R, Broil G, Pakula C, Jungbecker H, Braun M, Rüdenauer I, et al.
tor. In: Proceedings of the Vernetzungstagung, Berlin; 2010. Synergy potential of smart appliances. University of Bonn; 2008.
[61] Statistisches Jahrbuch. Red.-Schluss: 1. August 2011. Statistisches Bundesamt; [77] Minerals yearbookVolumes report international, vol. 3. United States
2011. Geological Survey; 2010.
[62] Haberkern B, Maier W, Schneider U. Steigerung der Energieeffizienz auf [78] European Environmental Agency. CORINE land cover raster data, version 13;
kommunalen Kläranlagen. Dessau-Roßlau: German Federal Environment 2010.
Agency (UBA); 2008. Report to the. [79] Gallego FJ. A population density grid of the European Union. Popul Environ
[63] Lanz M, Fricke B, Anthrakidis A, Genter M, Hoffschmidt B, Faber C, et al. CO2- 2009;31:460e73.
Emissionsminderung durch Ausbau, informationstechnische Vernetzung und [80] International Earth Science Information Network (CIESIN) CU Center for,
Netzoptimierung von Anlagen dezentraler, fluktuierender und erneuerbarer Agricultura Tropical (CIAT) CI deGridded population of the world, Version 3
Energienutzung in Deutschland. Dessau-Roßlau: German Federal Environ- (GPWv3); 2005.
ment Agency (UBA); 2011. Report to the. [81] Eurostat European Statistics Database. Section nrg_11: energy statistics e
[64] Dena. dena-Netzstudie II. Integration erneuerbarer Energien in die deutsche infractructure, dataset nrg_113a; as of November 26, 2013.
Stromversorgung im Zeitraum 2015e2020 mit Ausblick 2025. Deutsche [82] Pirker O, Argyrakis I, Babkin V, Chudy M, Crosnier G, Dahlback N, et al. Hydro
Energieagentur (dena); 2010. power in europe: powering renewables. Eurelectric; 2011.