d

Economic Benefits of Direct Current Technology for

we
Private Households and Peer-to-Peer Trading in Germany
Xueying Liu1 and Reinhard Madlener1,2 ∗

ie
1 Institute for Future Energy Consumer Needs and Behavior (FCN)/ E.ON Energy Research Center,
FEN Research Campus, School of Business and Economics, RWTH Aachen University, Aachen, Germany

ev
2 Department of Industrial Economics and Technology Management,
Norwegian University of Science and Technology (NTNU), 7491 Trondheim, Norway

r
With the increased adoption of solar photovoltaics (PV) and batteries, and the use of electronics
and appliances powered by direct current (DC), e.g. heat pumps, and electric vehicles (EVs), DC
er
technologies offer higher energy efficiency compared to the entrenched alternating current (AC)
technologies. However, the adoption of DC infrastructure is limited due to path dependency and
lock-in effects of the currently dominant electric infrastructure based on AC technology. Efficiency
pe
gains in energy communities and for households may facilitate the wider-scale adoption of DC
technologies. In this study, we simulate 600 household load profiles based on twelve different
representative household types and estimate the possible energy cost savings of a DC architecture
compared to an AC architecture under various electricity prices and feed-in-tariff levels. This is
done for different combinations of battery and PV sizes, and for the case of a peer-to-peer (P2P)
ot

trading community. The results show that the DC home yields cost savings of around €90 p.a. for
the median household when compared to an AC home. Moreover, we find that neither the share
tn

of DC load nor household characteristics impacts cost savings significantly, while the total load
remains the most important factor influencing the cost-saving potential. In addition, while cost
savings do not necessarily increase with larger PV and battery sizes, they do increase with the
possibility of households to engage in P2P trading. The results yield an improved understanding
rin

regarding the cost-saving potentials of DC homes and their expected diffusion in Germany. This
is especially relevant for future large-scale adoption of solar PV, batteries, and EVs in the future,
thus helping both policy-makers and companies alike to better assess the market potential of DC
ep

homes.

Keywords: DC technology; Choice of Technology; Diffusion; Industrial policy; Path dependence

JEL Classification Nos.: O14, O25, O33, O52
Pr

∗ Corresponding Author: Tel: +49 241 80 49820, RMadlener@eonerc.rwth-aachen.de

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 1 Introduction
Direct current (DC) technology is probably one of the oldest forms of electric technology, but

d
also has the potential to play an important role in the future of the electricity world. Recently,

we
new developments in the way that private households are changing their technological set-ups
and increasingly turning into prosumers, has made an evaluation of the economic promises
that DC technology holds for individual decision makers necessary. With the increase in solar
photovoltaic (PV) power generation and home batteries installed in private households, as well

ie
as the increasing use of consumer electronics, digital end user appliances, and other applications
that are powered by DC, e.g. heat pumps, electric vehicles (EVs), direct current technologies

ev
and transmission offer lower conversion losses and thus higher energy efficiency compared to
the currently entrenched alternating current (AC) technologies (Vossos et al., 2017).

r
In 2019 and 2020, German households installed around 136,000 new PV systems (Bun-
desnetzagentur, 2021b) and, in 2020, 70% were installed with a battery system (Bundesver-
er
band Energiespeicher, 2021). Computers and other digital devices have become commonplace
in almost every household over the last 20 years. Heat pumps have become the most-widely
pe
used heating technology for new builds in Germany in 2020 (Energiebilanzen, 2021) and will
continue to be the favored technology for the future (Bründlinger et al., 2018), and electric mo-
bility is touted to bring about the next big transition in individual transport, marching from
sales record to sales record worldwide. Thus, it is important for engineers and policy-makers
ot

alike to look at what the cost savings potential of private household DC architecture looks like
under these changing circumstances.
tn

DC technology, in addition, also offers further benefits, such as flexibility and simplicity of
design, improved reliability and safety, and potentially lower installation costs (Marchionini and
Zheng, 2018). However, the adoption and especially the retrofitting of DC infrastructure are
rin

strongly limited due to path dependency (Arthur, 1994) and the lock-in effect (Liebowitz and
Margolis, 1995) of the current electric infrastructure being predominantly built upon alternating
current (AC).
ep

However, local energy communities and the development of new city districts based upon DC
technology, sometimes referred to as “DC quarters”, are being increasingly discussed as a more
energy-efficient and also cost-efficient way to meet energy demands of those communities (De
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 Pascali and Bagaini, 2019). Those efficiency gains in energy communities and the development
of new urban districts can be easily based on DC technology and may offer an interesting

d
new prospect for the future of DC adoption among residential consumers and prosumers. In
addition, efficiency gains for individual households with a PV system and a battery will facilitate

we
the wider adoption of DC technology, as well as further standardization and economies of scale,
scope, and learning, which are expected to occur with a mass market diffusion of this technology.
We extend existing simulation models (e.g. by Glasgo et al., 2016 and Vossos et al., 2017)

ie
through the investigation of private household DC configurations, building upon the study
of Glasgo et al. (2016), and investigating the cost savings potential of DC infrastructure in

ev
homes in Germany under different scenarios. The remainder of this paper is divided into four
sections. Section 2 gives an overview of the current state of the literature with respect to the
economics of DC households. Section 3 introduces the methodology used, in particular the way

r
the simulation is set up and how cost savings are calculated. Section 4 presents the simulation
results obtained for the different set-ups studied. Section 5 concludes and offers a discussion of
er
the insights gained as well as some limitations of this study.
pe
2 Literature Review
DC technology, in principle, is a long-known technology, going back to Thomas Alva Edison
and contemporaries, but has seen increased interest and research activity recently, as techno-
ot

logical advances (e.g. power electronics) as well as changing regulatory and political conditions
render DC technology an ever more promising avenue for more sustainable and energy-efficient
tn

electricity supply systems.

There are several contributions in the literature that focus primarily on the technical via-
bility and performance of DC configurations. Overview studies, such as Garbesi et al. (2012),
rin

provide an extensive catalog of DC appliances and power systems, which informs choices in
our simulation, Sannino et al. (2002) examine the feasibility of DC networks in commercial
facilities, in particular a university department, and also explore the economic implications of
ep

their approach. Gerber et al. (2018) investigate the energy efficiency of AC and DC power
distribution networks in commercial buildings via simulation studies (based on Modelica), and
find that under the right circumstances DC buildings can achieve more than an 11% energy
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 efficiency gains over an alternative AC architecture, making DC technology seem particularly
suitable for zero net energy buildings and microgrids. In addition, Noritake et al. (2014) re-

d
ported on a demonstrative research on DC microgrid solutions for office buildings in an energy
community microgrid setting in Japan, they find in their study that DC technology increases

we
energy efficiency and decreases electricity use from the grid by 4.2%. Weiss et al. (2015) present
a study on a DC office building prototype realized in Germany. The authors find overall system
energy efficiency gains of about 5%, which increase when combined with local PV generation,

ie
and they describe the general benefits of a low-voltage DC grid in similar settings.
Moreover, some general studies include both a technical and an economic aspect, such as

ev
Backhaus et al. (2015) and Johnson et al. (2016), which are scoping studies on DC microgrids in
which the authors find that microgrids, DC distribution systems, and on-site PV generation in
combination can offer benefits which are greater than the sums of the benefits of the individual

r
parts. Mackay et al. (2017) give a qualitative assessment of a universal DC distribution system
including DC grid architecture, electricity market design, control, and protection as well as
er
possible future use cases, such as microgrid and offgrid systems. They find that for many of
these use cases, as well as for distribution grids overall, a convincing case can be made for
pe
the use of DC technology due to increased energy efficiency and flexibility. In addition, Ullah
et al. (2020) examine the technical and financial benefit of a DC microgrid compared to an AC
microgrid for different PV and load types (e.g. PV with DC and AC load) and find that load
and system stability during fault condition are important factors in the wide adaptation of the
ot

DC system.
While many of the above-mentioned studies adopt simulation-based methodologies, there
tn

are also contributions to the literature that employ expert surveys, such as Vossos et al. (2017),
who give a general overview of DC technologies and the market situations by providing over
200 references to studies and demonstration projects, and Marchionini and Zheng (2018) who
rin

present an expert assessment of the potentials for the use of DC technology in buildings with
respect to the market benefits of DC, market opportunities and various barriers. They identify
improved energy efficiency as the most agreed-upon benefits of DC technology in buildings, in-
ep

dustrial use with the least opportunity, and existing AC infrastructure as the biggest challenge.
In addition, while many of the above-mentioned studies investigate DC buildings in general,
there are a limited number of studies which focus on household DC configurations. For instance,
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 a seminal contribution by Glasgo et al. (2016) examines the energy and cost savings potential
of DC houses in Texas, focusing on the difference in energy and cost savings via the adoption

d
of various DC appliances (e.g. DC refrigerator, DC condensing unit, EV). They find, given
real-world load data and solar profiles, that DC-based distributed electricity generated from

we
local PV can generate energy savings, whereas net savings adjusted for investment costs from
DC were currently not cost-effective at the time of the study. However, they also point out
that several non-technical barriers need to be overcome before DC distribution technology can

ie
be widely adopted in residential settings in the United States. The non-technical difficulties
that DC technology needs to overcome are identified as a lack of standardization, a relatively

ev
small market (and thus currently limited applications for DC components), as well as a lack of
training and experience of electrical workers with DC hardware. Whilst some of these hurdles
still remain today, it is likely that, for example, the limited availability of components will

r
not persist, so that new insights into the economic viability of DC-distributed, PV-generating
prosumer homes are to be expected from more recent studies, which is exactly what the study
er
at hand is aiming to achieve, albeit through the study of a different local market with its own
regulatory characteristics. Vossos et al. (2014) also look at the energy cost savings of residential
pe
buildings in 14 cities in the U.S., focusing on the effect of climate and battery storage; they
find an energy efficiency gain of 5% for DC buildings, and of even up to 14% for DC-powered
buildings which also make use of a home battery storage. Rasheed et al. (2019) study the
efficiency gains under different power electronic converter typologies, finding energy efficiency
ot

gains for DC in a similar range as other studies (i.e. ≈ 4%).

However, most of the above literature adopt a setup with fixed prices and economic analysis
tn

focused mainly on the US context, whereas other geographical regions are still very much
underrepresented. There are also relatively few policy implications to be found in the literature,
since the focus of many studies is mainly on technological and engineering viability, due to
rin

DC household technology having been in the prototyping phase until very recently. Studies
often show some economic implications, but with few exceptions the focus has been on energy
efficiency rather than cost efficiency and/or cost effectiveness.
ep

Furthermore, P2P trading and other concepts developed for energy communities are hardly
ever considered in a DC household context, despite such communities playing a significant part
in the discussion of future decentralized energy supply.
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 As such, we contribute to the existing literature by focusing on the monetary savings than
can be achieved by using a DC household architecture - in comparison to an AC household

d
architecture - and, furthermore, our setup is considering German households, bringing into
the debate a different load profile, PV generation profile, and regulatory framework. We also

we
investigate various scenarios and variations including household characteristics, the share of DC
load, different feed-in tariff (FIT) and electricity prices, and PV and battery size, to gain holistic
insights that are generalizable to a wide range of real-world private households. Furthermore, as

ie
the topic of P2P trading is gaining momentum, and a regulatory framework has been defined for
these business models, we also include a scenario analysis for a DC-based P2P energy trading

ev
community compared to an AC-based P2P energy community. Lastly, policy implications are
derived based on the results from the above-mentioned scenario studies.

r
3 Methodology and Data er
In this study, we simulate household load profiles based on twelve different representative house-
hold types and estimate the possible cost savings achievable by means of a DC grid configuration
pe
compared to an AC system. The simulation considers different possible combinations of PV
and batteries (both varying in size) for each simulated household type. The annual cost sav-
ings are calculated for different electricity price development scenarios. Households with the
highest expected cost savings are also identified. A major assumption of this study is that the
ot

households’ decisions to adopt solar PV and batteries is independent of the adoption of DC

circuit infrastructure. In future studies, this assumption could be relaxed and the decisions of
tn

private households could be modeled as simultaneous. The entire model set-up is illustrated
by Figure 1. Details of the methods are provided in the following sections.
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 d
ie we
r ev
Figure 1: Overview of Model Set-up
er
pe
3.1 AC and DC Architecture and Efficiency Calculation

This section details the design of AC and DC system configurations in private households as well
as in the DC neighborhood that allow for P2P trading. Many different system configurations
exist, described for instance in Vossos et al. (2014), Backhaus et al. (2015), Glasgo et al.
ot

(2016), and Gerber et al. (2018) among others. We used a combination of the above to arrive
at configurations that are up-to-date, efficient for both DC and AC, and for which there are
tn

available data in terms of energy efficiency. The configurations analyzed are shown in Figure 2.
The highest possible energy efficiency level of the best available technology for each component
is taken from the literature to ensure comparability, and as a forward-looking measure, since we
rin

assume future improvement in energy efficiencies. As a result, it may appear that the efficiencies
here are higher than those typically achieved currently in many real-world applications.
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 d
we
(a) AC (b) DC

ie
Figure 2: AC and DC Residential Home Configurations

ev
In addition, we also consider the case of a peer-to-peer (P2P) energy-trading community.
Based on the German “Direktlieferung” framework (NRW, 2021), P2P trading allows the saving
of grid fees and electricity taxes. However, it also requires that no public grid is used (NRW,

r
2021). As a result, there is potential for an energy-sharing community to adopt DC grid
er
configuration. The grid configuration considered is shown in Figure 3.
pe
ot
tn
rin

Figure 3: DC Residential Home Configuration for an Energy Sharing Community

ep

Based on the above-shown configurations and energy efficiencies, the system energy efficien-
cies calculated are shown in Table 1.
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 Table 1: System Energy Efficiencies for Different Residential Power Supply Configura-
tions

Scenarios AC DC DC Community

d
PV to DC Load 90.8% 97.5% 97.5%
PV to AC Load 91.7% 98.5% 98.5%

we
Battery to DC Load 87.6% 97.0% 97.0%
Battery to AC Load 88.5% 98.0% 98.0%
PV to Battery 91.7% 98.5% 98.5%
Grid to DC Load 93.1% 96.0% 98.0%

ie
Grid to AC Load 94.1% 97.0% 99.0%
PV to Grid 97.5% 97.5% 99.5%

ev
3.2 Simulation of Household Load Profile

r
We use the load profile generator developed by Pflugradt (2017) to simulate the load profile
of the 12 household types shown in Table 2. The average annual simulated loads for each
er
household type are also shown in Table 2. We simulated hourly load profiles, and for each
household type, we simulated 50 households. Therefore, we have 600 households in the dataset
pe
and 5,256,000 data points. PV and battery are not included in these simulations, but are added
later on in order to be able to investigate scenarios with different PV and battery sizes.
Note that for two reasons we did not include an electric vehicle in our current simulations:
(1) it is unsure whether the charging will take place mainly at home or at the workplace and
ot

other charging stations; and (2) we do not want to make assumptions regarding the adoption
of electric vehicles. However, our results do show that electric vehicle charging generally would
tn

increase the cost savings of a DC system, since the cost savings increase linearly with energy
consumption and do not vary significantly by load profile (see subsection 4.2).
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 Table 2: Types of Households Considered and Annual Loads Assumed

Size Work Status Average Annual Load (kWh)

Single, No Child Working 2419

d
Single, No Child Retired 3593

we
Single Parent, 1 Child Working 3059
Single Parent, 2 or More Children Working 4049
Couple, No Child Working 4314
Couple, No Child Retired 5983

ie
Couple, 1 Child Both Working 4505
Couple, 1 Child Single-Earner Family 4424
Couple, 2 Children Both Working 5591

ev
Couple, 2 Children Single-Earner Family 5768
Couple, 3 or More Children Both Working 5778
Couple, 3 or More Children Single-Earner Family 5775

r
3.3
er
Calculation of Energy Generation, Consumption, and Sharing

We use an hourly resolution for all calculations. For energy generation by PV, we apply capacity
pe
factor data from Pfenninger and Staffell (2016) and take the average for each hour in the year
from 2000 to 2019 for Germany. We use the capacity factor, since we would like to vary the PV
size and see its impact on the overall results. The maximum battery depth of discharge (DiD)
is set to 90% (Klingler, 2017).
ot

As the electricity price is much higher than the FIT in Germany, self-consumption is pre-
ferred (Beck et al., 2016). This means that the PV generated electricity is self-consumed first,
tn

the remaining is stored in the battery and, once the battery is full, the rest is fed into the grid.
Vice versa, when the generated electricity is not enough to supply self-consumption, energy
is taken first from the battery. Once the battery is empty, electricity is drawn from the grid.
rin

For energy sharing, we assume the best-case scenario for each household, which means that
all the electricity generated from PV is shared within the community rather than fed into the
the grid. In the case where generated electricity is not enough for self-consumption, all the
ep

energy is taken from the shared community. This might not be realistic; however, it represents
a best-case scenario and we then know that the results obtained indicate an upper bound in
terms of potential energy cost savings to be achieved, which is useful information in itself.
Pr

For each step of the energy flow, we applied the efficiency values presented in Table 1.

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 Mathematically, the above rule can be formulated as:

d

min[SoC t−1 + pvt − loadt , SoCmax ] if pvt − loadt > 0


SoCt = (1)

we
t−1 − (loadt − pvt ), otherwise.

max[SoC

DiD]

ie

0 if pvt − loadt > 0


f rom grid = (2)
t − pvt − (SoCt−1 − SoCt ) otherwise.

load


ev

pv t − loadt − (SoCt − SoCt−1 ) if pvt − loadt > 0



r
f eed in = (3)
0 otherwise.


er
SoCt represents the state of charge of the battery after period t, and it is charged until full
charge SoCmax is reached if the electricity generated by PV, pvt , more than supplies for the
pe
load at the time, loadt . If pvt is not enough to supply the own load, the battery is discharged
until a pre-set maximum depth of discharge (DiD, set to 90% based on Klingler (2017))is
reached. Energy is drawn from the grid, f rom grid, if battery and PV generated electricity is
insufficient. Electricity is fed into the grid, f eed in, if there is any excess after own consumption
ot

and battery charging.

tn

3.4 Investment Costs

To evaluate the decision of the adoption of DC configurations, it is important to examine the

investment costs as well. The investment costs here refer to the upfront costs of installing the
rin

system, including for instance line equipment and converter equipment. Since DC configurations
and technologies are not fully commercialized yet, there is a lack of data for DC appliances,
ep

converters, and system installation costs (Vossos et al., 2017), we take estimates from existing
literature.
For new build, Backhaus et al. (2015) estimated that the cost of building a DC micro-grid is
Pr

half of its AC counterpart. Moreover, Vossos et al. (2017) stated that DC systems in buildings

10

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 have the potential to have a lower investment cost compared to AC systems in buildings, mainly
due to the simplicity of the circuitry and fewer components used. The study suggests that DC

d
converter costs should be less than the AC counterpart in principle, and DC systems require
fewer components, hence lower manufacturing and consumer costs. However, Glasgo et al.

we
(2016) estimated additional annualized investment and operating costs of a DC home to be
$380 for bidirectional inverter on average and $900 for DC appliances and component (€323 to
€765, respectively, at an exchange rate of €0.85 to $1), while pointing out that the additional

ie
costs of DC appliances and components could be eliminated due to widespread deployment
over time.

ev
For retrofit, (Glasgo et al., 2016) estimated LAC of $6000 to $10,000 (€5100 to €8500 at an
exchange rate of €0.85 to $1 ) for homes and stated that the retrofit costs are still prohibitively
high for DC systems. In addition, due to the lack of application of DC systems so far, it is

r
difficult to estimate other important installation and system operation and maintenance costs,
such as labor costs and insurance premia required for DC homes.
er
4 Simulation Results and Policy Implications
pe
The simulation results are presented next. Note that we do not consider the upfront investment
costs in the simulations, but only focus on the energy cost savings due to increased efficiency
from DC configurations. This separation allows us to focus on the return on investment, i.e.
ot

cost savings, which can be compared with investment costs detailed in section 3.4 or with
investment costs in other studies. We ran a first standard simulation to obtain an overview of
tn

the cost saving potential by DC configuration compared to AC. For this base case scenario, we
use the optimal PV and battery size calculation method based on the long-term scenario given
in Weniger et al. (2014), since that study is specifically customized to the situation in Germany.
rin

Consequently, we also used the electricity price and FIT from their study for consistency and
direct comparisons. The results are presented in Figure 4. Note that the annual cost savings
are the difference in cost savings between DC and AC configurations. We observe that the
ep

majority of households (around 30%) save around €80 additional per annum by adopting a DC
configuration. On the one end of the spectrum, some households only save around €20, while
on the other end, about 2% of households manage to save around €140 per annum.
Pr

11

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 d
ie we
Figure 4: Histogram of Households’ Annual Cost Savings from DC (Optimal PV and Battery

ev
Size)

r
In addition, we ran scenario simulations to answer the following five questions:
(1) How does the percentage of DC load vs AC load affect the cost savings potential of the DC
er
configuration?
(2) How does the household type affect the cost savings potential of the DC configuration?
pe
(3) How do different PV and battery size combinations affect the savings potential of the DC
configuration?
(4) How do FIT and electricity price affect the cost savings potential of the DC configuration?
(5) How does P2P trading affect the cost savings potential of DC configuration?
ot

For simulations (1), (2), and (5), we used the optimal PV and battery size calculation
tn

method introduced in Weniger et al. (2014) as well as the price given (FIT at 0.02 €/kWh
and electricity price at 0.34 €/kWh). For simulation (3), since we can vary the size of PV and
battery, we do not need to follow the values given in Weniger et al. (2014). Hence we simulated
rin

the results based on two sets of electricity price and FIT values, namely, current price level
and forecasted future price level. We use the official data for the current electricity price level
given by Destatis (2021) and the FIT given by Bundesnetzagentur (2021a). For future price
ep

levels, we use the predicted long-term values reported in Klingler (2017) (FIT at 0 €/kWh
and an electricity price of 0.266 €/kWh), since it is more recent than Weniger et al. (2014).
For simulation (4), we again used the aforementioned optimal PV and battery size calculation
Pr

12

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 method and the maximum and minimum values found in the literature and official websites
as a lower and an upper bound for the electricity price and FIT, and we simulate scenarios in

d
between the bounds.

we
Table 3: Values Used for the Price and FIT Simulation

Parameter Value Source

FIT Lower Bound (€/kWh) 0 Klingler (2017)

ie
FIT Higher Bound (€/kWh) 0.08 Rounded Value of Bundesnetzagentur (2021a)
Electricity Price, Lower Bound (€/kWh) 0.26 Rounded Value of Klingler (2017)
Electricity Price, Upper Bound (€/kWh) 0.34 Weniger et al. (2014)

ev
The results in this section should ideally also inform policy-makers in designing potential
policies that best target private households with the highest potential cost savings achievable

r
in houses. Policy-makers may see the potential to promote DC systems for two reasons. Firstly,
er
there are considerable potential for greenhouse gases mitigation, efficiency gains, and energy
saving gains in DC systems based on Life Cycle Assessment (Kockel et al., 2020).
Moreover, based on the investment costs for new build (see section 3.4) and the annual
pe
cost savings found in this study, currently no households have an economic incentive to adopt
a DC system without public policy financial support. Since investment costs are expected to
fall and become even lower than AC counterparts with a wide-spread adoption and diffusion of
ot

DC technology due to the simplicity of the circuitry, fewer components, and lower costs of DC
converters (see section 3.4), once policy support has kick-started the mass market diffusion of
DC applications, DC technology diffusion may become self-perpetuating.
tn

The results for the above mentioned five sets of simulations are presented in the following
section. In each subsection, we first briefly point out the simulation setup. Afterwards, we
present the simulation results and the interpretation of the results. Finally, we present some
rin

policy implications based on the results and our interpretation.

ep

4.1 Percentage of DC Load

For this simulation, we use the optimal PV and battery size, and vary the share of DC load in
total load from 0 to 1 for each household type considered. The results in terms of the mean
Pr

annual cost savings for each household type are plotted in Figure 5.

13

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 d
ie we
r ev
er
Figure 5: Annual Cost Savings from DC Configuration vs Share of DC Load (Optimal PV and
pe
Battery Size)

From Figure 5, we observe that the cost savings from DC configuration do increase with
ot

higher DC load, since the conversion is more energy efficient. However, the increase is marginal
and suggests that load type does not influence energy efficiency significantly. Three policy
tn

insights can be gathered from this observation. Firstly, DC-promoting policies should focus
on the power generation and energy storage technologies, e.g. PV and battery, rather than on
load type. Secondly, the cost savings alone from improved energy efficiency are not attractive
rin

enough to incentivize the adoption of DC configurations. Therefore, there is a need to further

decrease the investment costs associated with DC configurations and to advertise this once
costs have been brought down (see section 3.4 for more details).
ep

As the results here show that the percentage share of the DC load does not significantly
change the cost savings, for all the simulations presented in the following, we use a 50% DC
load and a 50% AC load assumption for simplicity reasons and do not vary the share of DC
Pr

load anymore.

14

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 4.2 Household Types Considered

For this simulation, we use the household types defined in Table 2, the optimal PV and battery

d
sizing methods, as well as the prices reported in Weniger et al. (2014). The results for all

we
600 households are plotted in Figure 6. The savings refer to the cost savings due to the
adoption of DC configurations alone, since we compare the system costs with an equivalent
AC configuration. The colors represent different household types. The results show that the
savings potential depends mainly on the annual total load, and is rather insensitive to the

ie
individual household’s load profile shape or household type. The main reason that the figure
shows clusters based on the household type - e.g. households with singles are clustered around

ev
the bottom left corner and couples with three children on the top right corner - is due to the
fact that households of the same type usually have similar total annual loads. Based on the
results, we also calculated that for each additional kWh load in a year, the cost savings increase

r
by €0.0168. er
pe
ot
tn
rin
ep

Figure 6: Annual Cost Savings from DC Configuration vs Household Loads (Optimal PV and
Battery Size)
Pr

15

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 The above results have three policy implications. First of all, they suggest that DC-
promoting policies should ideally target heavy electricity users. This means that with elec-

d
trification and adoption of more heat pumps and electric vehicles, it becomes increasingly
attractive to adopt DC configurations. In addition, it also suggests that any DC-promotion

we
policy should consider building types with heavier loads as well, e.g., multifamily houses, com-
mercial buildings. Secondly, it also gives families and policy planners a rule-of-thumb value,
namely 0.0168 €/kWh, to estimate what additional cost savings can be achieved through ad-

ie
ditional load, e.g. an electric vehicle. This value appears rather low at the moment, especially
given that the current FIT is at 0.078 €/kWh (Bundesnetzagentur, 2021a). However, the FIT

ev
is estimated to go to nearly 0 in 2040, based on Weniger et al. (2014), on Klingler (2017),
and our own estimations following the German Renewable Energy Act (Art. 49 EEG 2021),
assuming that PV installation always meets the set target. These additional savings from using

r
DC technology could become more attractive then, thus providing a reasoning for promoting
DC configurations. Thirdly, it suggests that the diffusion of electric vehicles could facilitate the
er
diffusion of DC configurations due to increased load.
pe
4.3 PV and Battery Size

For this simulation, we vary the size of PV and battery and assume that when no PV is installed,
families will also not install a battery. The PV sizes are set from 0 kWp to 10 kWp , and the
ot

battery sizes from 0 kWh to 10 kWh. We simulated for the current price level and for future
price levels, based on values introduced at the beginning of section 4. The results are plotted
tn

in Figure 7. The cost savings refer to the average savings across all 600 households considered.
For both sets of prices, the following observations can be made:
(1) Savings are maximized at a 2.5 kWp PV and a 10 kWh battery size.
rin

(2) For larger PV sizes, cost savings are maximized for a battery size of 7.5 kWh.
(3) Cost savings do not increase with increasing PV size.
ep
Pr

16

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 d
ie we
(a) Current Price (b) Expected Future Price

Figure 7: Annual Average Cost Savings for Different PV and Battery Sizes

ev
These findings seems counter-intuitive at a first glance. For a better understanding, mean
cost savings from an AC configuration and a DC configuration individually are presented in

r
Figure 8. It can be observed that the annual system cost savings increase with increasing PV
er
and battery size for both configurations (note that we do not consider the investment costs here,
but only the electricity costs and gains from FIT). This follows our expectation since for large
PV size, households can generate more electricity for self-consumption and for gird feed-in,
pe
resulting in lower electricity bill and higher gains from FIT. Similarly, with larger battery sizes,
more self-consumption is possible, resulting in lower electricity bills. The savings from the DC
configuration compared to an equivalent AC system shown in Figure 7 reveal the difference
ot

between the two configurations. The findings are not as straightforward as the expectation
that cost savings would increase with larger PV and battery size.
tn
rin
ep

(a) AC (b) DC
Pr

Figure 8: Total System Cost Savings from the DC v.s. AC Configuration

17

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 The observations from above suggest that the cost savings from the adoption of a DC
system are rather modest when no battery is installed. Consequently, DC promotion policy

d
should ideally promote battery adoption at the same time or should be launched when battery
adoption is picking up and in areas with battery promoting policies. Moreover, since the

we
cost savings do not increase with PV size, DC-promoting policies do not need to focus on the
adoption of larger PV systems. In addition, for each PV system size, there is an optimal battery
size that maximizes the cost savings from the DC configuration. However, this size combination

ie
does not necessarily correspond to the optimal PV and battery size for households if overall
cost savings, including investment costs, are to be maximized. As such, DC-promoting policies

ev
should strike a balance between the two different optimal sizing methods, which might not be
that easy in practice, also due to the dynamics involved.

r
4.4 Electricity Price and FIT Levels
er
For this simulation, we vary the electricity price and value of the revenues from the FIT. The
lower and upper bounds for our simulations are detailed in Table 3. The PV and battery
sizes are set to be optimal based on the aforementioned methods. To better illustrate the cost
pe
savings potential, we plotted both the DC adoption cost savings compared to an equivalent AC
system, as well as compared to an AC system with no PV system or battery, shown in Figure
9.
ot
tn
rin
ep

(a) DC Compared to Equivalent AC System (b) DC Compared to AC System w/o PV/Battery

Figure 9: Annual Cost Savings for Different Electricity Prices and FIT Levels (Optimal PV
and Battery Size)
Pr

18

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 It can be observed from Figure 9 that the cost savings increase with increasing electricity
price and FIT levels. This is expected, since the DC configuration is generally more energy

d
efficient, and for the same amount of energy saved, the higher the electricity price and the
FIT level are, the higher the monetary values associated with the amount of saved energy. In

we
addition, the comparison between the two figures side by side reveals that the magnitude of
the annual cost savings from DC configuration is rather small compared to the savings due to
the adoption of PV system and a battery only.

ie
The results suggest that savings will decrease, as the FIT is lowered regularly based on
the German Renewable Energies Act (§ 49 EEG 2021) and with the expected peaking of the

ev
electricity price around 2030 (Schlesinger et al., 2014).

4.5 Peer-to-Peer Trading

r
As detailed in section 3.3, we assume that for each household, only the P2P trading price
er
applies (hence the ideal case where the buying price is lower than the buying price from the
public grid and the selling price is higher than the FIT). We adopt the decentralized battery
storage setup based on Lüth et al. (2018) where the P2P price is the same for the seller and
pe
the buyer, i.e. the amount that one of them pays is the same as the amount that the other
one earns. While the actual price is dependent on the design of the trading scheme and the
negotiation between members, for our simulation, we follow the approach used in Zepter et
ot

al. (2019) and adapt it to the case of Germany. Zepter et al. (2019) state that the P2P price
should lie between the feed-in-tariff and the price of electricity from the grid in order to promote
tn

local trade, and a reasonable price would be the price of the electricity from the grid minus
the transmission network charges. In Germany, P2P trading does not need to pay electricity
taxes or grid fees (NRW, 2021), which together amounted to 28.7% of the total electricity
rin

price in 2020 (Bundesnetzagentur, 2021c); (the total electricity is 32.05 €ct/kWh, of which 7.14
€ct/kWh are grid fees and 2.05 €ct/kWh are electricity taxes). Therefore, for our calculation,
we assume that the P2P price will be 71.3% of the price for the grid electricity.
ep

The results, in comparison with the no P2P trading case, are plotted in Figure 10, as well
as the results for scenarios with different grid electricity prices. Note that the cost savings here
still refer to the net savings when compared with an equivalent AC system with P2P. When
Pr

comparing to a setup without P2P trading, P2P trading results in two types of additional cost

19

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 savings for DC configuration: (1) cost savings from trading i.e. lower buying price and higher
selling price, (2) additional energy efficiency gains due to DC technology. Note that the first

d
type of cost savings are deliberately excluded from the figures, since an equivalent AC system
would also allow for this type of cost savings.

ie we
r ev
Figure 10: Histogram of Household Annual Cost Savings with a DC system, with and without
P2P trading (Optimal PV and Battery Size)
er
pe
ot
tn
rin

Figure 11: Average Annual Cost Savings from DC Configuration with P2P Trading for Different
Electricity Price Scenarios (Optimal PV and Battery Size)
ep

From the histogram in Figure 10, we can observe that the overall cost savings increase due
to P2P trading. As shown in Figure 9, savings from DC increase with higher FIT and higher
electricity price levels, i.e. cost savings increase with higher selling prices and higher buying
Pr

20

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 prices of electricity. It can hence be deduced here that, due to P2P trading, the effect of the
increase in the selling price compared to the FIT level outweighs the decrease in the buying

d
price compared to the grid electricity price, leading to an overall increase in cost savings.
Based on this observation, two policy implications can be derived. Firstly, any DC household

we
pilot program should ideally be designed with the flexibility of allowing P2P trading in order
to increase cost savings. Secondly, DC configuration diffusion is more likely to start after P2P
trading has been more widely adopted. Section 3.4 explains that the upfront investment costs

ie
are prohibitively high for retrofits but have the potential to be even lower than AC systems
with new build. Based on the German P2P trading framework “Direktlieferung”, new private

ev
electric grids need to be built locally anyway. As such, with higher cost savings potential, the
need to build anew, and potential lower investment costs, a DC configuration is more likely to
start diffusing into the market together with P2P trading. This insight could inform the timing

r
of the introduction of DC promotion technologies.
er
5 Conclusion
pe
Our study has certain limitations that should be noted. First of all, as mentioned in the model
setup, our study is limited to private households in Germany, as we use simulated load and
generation profiles for Germany, as well as the German regulatory framework with respect to
FIT, P2P trading regulations, etc. However, it should also be noted that our methodology can
ot

easily be adapted to other geographical and regulatory contexts if sufficient data are available.
The size of the effects found may vary based on different contexts, but the overall comparison
tn

between AC and DC architectures should still yield comparable results. Our results show that
the DC cost-saving potential increases with higher loads and that the load is the main factor
influencing the cost savings potential. This suggests, on the one hand, that we should take other
rin

energy-intensive applications with higher load into consideration e.g. commercial or industrial
sites. On the other hand, the effects become more pronounced for households that will exhibit
a higher load in the future, e.g. households that are switching from traditional fossil-fuel-based
ep

home heating solutions to heat pumps, or households increasingly adopting EVs over traditional
internal combustion-engine vehicles (see below).
Secondly, our study uses simulated rather than actual data for the load profile and PV
Pr

21

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 generation. Whilst the profile generators used are generally considered to be quite reliable and
their estimates to be quite robust, it would of course be sensible to cross-check our results with

d
real-world data - both to corroborate our findings and to corroborate the reliability of the load
profile generators for future research. Similarly, when simulating the battery charging pattern,

we
we did so by using an optimizing algorithm, rather than the actual charging behavior of DC
households. These two aspects are done this way because DC households are still very rare and
no other data are currently available to us. However, before any large-scale deployment, more

ie
pilot projects should ideally be carried out to generate actual data ond load and generation
profiles and charging behavior for another study, in order to better observe the use of DC in

ev
reality. Nevertheless, we find the use of an optimizing algorithm prudent given the current data
limitations, as we expect to see optimizing algorithms governing the charging behavior of home
batteries through smart home automatization in the future such that our relatively simplistic

r
algorithm is an adequate heuristic given the limited options of behavior in our model.
Thirdly, we did not include EVs in our study, which we expect to be increasingly adopted
er
by private households. The effect of owning and operating an EV is indirectly shown by the
results, since the addition of an EV generally leads to higher load and thus results in higher
pe
monetary savings when compared to an AC system. As such, we deliberately choose to exclude
EVs from this study because we did not want to make more assumptions than necessary with
respect to the percentage of households which will adopt an EV at a certain point in time,
nor on their use and charging patterns. We expect that under some reasonable assumptions,
ot

the results would be even more favorable to the DC configuration, but this would be entirely
assumption-driven. Once the adoption pattern and charging behavior of EVs is more clear, they
tn

could be included within the analytical framework of our study. Therefore, with the increased
rate of EV adoption, i.e. when fewer assumptions are needed due to better data availability,
a study comparing DC with AC involving EVs directly as well should be conducted with our
rin

framework or a modified version thereof.

Fourthly, we focus on the annual cost savings due to improved energy efficiency of DC
configurations, but did not investigate the investment costs. Some investment cost figures can
ep

be gathered from the literature, but we have some reservations about their usefulness over the
medium and long term because these cost estimates are strongly influenced by technological
development and engineering, insurance premia, market structure and the level of competition,
Pr

22

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 and the future market price that firms will charge to provide such DC solutions to households.
These are all factors that are largely still in development and not yet commercially available

d
to private households. We expect, however, that once learning curve effects of production
become effective, as well as economies of scale are made possible due to a wider uptake of DC

we
technology, that the currently relatively high investment costs for DC technology will come
down significantly. Once this market is more developed and mature, and thus more data also
become available, a comparison between the investment costs and the annual cost savings should

ie
be done to gauge the overall economic viability of DC solutions. Another factor that needs to
be considered with respect to investment costs is the competitiveness of the DC appliance and

ev
circuitry market, but this is beyond the scope of the present study and eventually needs to be
investigated in its own framework of industrial organization.
Lastly, we assumed the best-case scenarios for P2P trading to simplify our calculation

r
process. Whilst this makes sense at the current state-of-the-art of P2P trading, more studies
could be conducted when P2P platforms have matured more and actual data from P2P trading
have become available.
er
The results show that the energy efficiency improvement of DC household architecture yields
pe
cost savings of around €90 p.a. for the median household when compared to a traditional AC
architecture. Whilst this is non-negligible with respect to the average household electricity
bill i.e. annual average at €1065 in 2019 according to BDEW (2019), this might also not be
sufficient to induce a large number of households to move away from the currently widespread
ot

and true-and-tested AC technology architecture in private households. This suggests the need
to reduce investment costs further, which currently exceed those annual cost savings, to make
tn

DC more attractive. Especially retrofitting households with a working AC architecture seems

economically nonviable in the foreseeable future, as many one-time costs for construction and
remodeling are involved. We do, however, expect to see economic viability for retrofitting DC
rin

architecture when the wiring of a household is scheduled for maintenance and updating anyway,
e.g., after the end of the expected lifetime of the currently used circuitry or when updating
old circuit layouts to accommodate more modern home needs, e.g. more outlets or a different
ep

circuit topology for smart home applications.

Moreover, we found that the share of DC load does not impact cost savings significantly, and
the focus should therefore be on generation and storage. Similarly, household characteristics do
Pr

23

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 not impact cost savings significantly either, and still the total load remains the most important
factor influencing the cost saving potential.

d
Lastly, while cost savings do not necessarily increase with larger PV and battery sizes,
they do increase with the possibility of households to engage in P2P trading, furthermore

we
emphasizing the importance that P2P trading in energy communities can have for achieving
cost efficiency in a decentralized electricity system. Policies aimed at promoting efficiency
through DC architecture should thus also consider supporting energy communities and P2P

ie
trading.
Overall, our results yield an improved understanding regarding the achievable cost savings

ev
of DC homes under different scenarios and inform on the potential diffusion of DC homes
in Germany. That country, given its widespread adoption and subsidization of PV systems,
provides meaningful insights into the future diffusion of DC household architectures on the

r
international level, even though one needs to be careful when generalizing the results to other
countries, as the absolute cost savings potential might be biased slightly upwards for the German
er
case due to the relatively high electricity prices (including fees and taxes) in an international
comparison. However, we expect an increased reliance on self-generated electricity and less
pe
reliance on the grid to mitigate these effects at least somewhat. This study is especially relevant
as we expect a large-scale adoption of solar PV, battery and EVs in the future by private
households. Our insights on the savings potential and overall load, as the primary determining
factor thereof, can be helping policy-makers, on a national and local level alike, urban planners,
ot

as well as companies to better assess the potential of DC household architectures, which we

expect to be especially relevant with respect to new developments, where there is no cost of
tn

retrofitting, but where energy efficiency gains and cost savings can be realized.
rin

Acknowledgments
We would like to acknowledge the fruitful discussions with colleagues from the Research Campus
FEN at RWTH Aachen University (www.fen.rwth-aachen.de). This research was funded by
ep

the German Federal Ministry of Education and Research (BMBF, FKZ 03SF0592), Flexible
Electrical Networks (FEN) Research Campus.
Pr

24

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 References
Arthur, W Brian et al. (1994). Increasing returns and path dependence in the economy. Univer-

d
sity of Michigan Press.

we
Backhaus, Scott N, Gregory William Swift, Spyridon Chatzivasileiadis, William Tschudi, Steven
Glover, Michael Starke, Jianhui Wang, Meng Yue, and Donald Hammerstrom (2015). DC
microgrids scoping study. Estimate of technical and economic benefits. Tech. rep. Los Alamos
National Lab.(LANL), Los Alamos, NM (United States).

ie
BDEW, Bundesverband der Energie und Wasserwirtschaft (2019). Stromrechnung für Haushalte.

ev
url: https://www.bdew.de/media/documents/20190723_BDEW_Entwicklung_Preisbes
tandteile_Strom.pdf (visited on 06/18/2021).
Beck, Tobias, Hendrik Kondziella, Guillaume Huard, and Thomas Bruckner (2016). “Assessing

r
the influence of the temporal resolution of electrical load and PV generation profiles on
self-consumption and sizing of PV-battery systems”. Applied energy 173, pp. 331–342.
er
Bründlinger, Thomas, JE König, O Frank, D Gründig, C Jugel, P Kraft, O Krieger, S Mischinger,
P Prein, H Seidl, et al. (2018). “dena-Leitstudie Integrierte Energiewende: Impulse für die
pe
Gestaltung des Energiesystems bis 2050”. Deutsche Energie-Agentur GmbH (dena), ewi
Energy Research & Scenarios gGmbH: Berlin/Köln, Germany.
Bundesnetzagentur (2021a). Degression rates and reference values for payment February - April
2021. url: https://www.bundesnetzagentur.de/EN/Areas/Energy/Companies/Renewab
ot

leEnergy/Facts_Figures_EEG/Register_data_tariffs/EEG_registerdata_payments_
node.html (visited on 05/27/2021).
tn

— (2021b). Gesamtdatenauszug der öffentlichen Daten des Marktstammdatenregisters. url: ht

tps://www.marktstammdatenregister.de/MaStR/Datendownload (visited on 06/18/2021).
Bundesnetzagentur, Bundeskartellamtes (2021c). Monitoringbericht 2020 der Bundesnetzagen-
rin

tur und des Bundeskartellamtes. url: https://www.bundesnetzagentur.de/SharedDocs/

Mediathek/Berichte/2020/Monitoringbericht_Energie2020.pdf?__blob=publicatio
nFile&v=8 (visited on 05/18/2021).
ep

Bundesverband Energiespeicher, BVES (2021). BVES Branchenanalyse 2021 - Entwicklung und

Perspektiven der Energiespeicherbranche in Deutschland. url: https://www.bves.de/wp-
content/uploads/2021/03/2021_BVES_Branchenanalyse.pdf (visited on 06/18/2021).
Pr

25

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 De Pascali, Paolo and Annamaria Bagaini (2019). “Energy Transition and Urban Planning for
Local Development. A Critical Review of the Evolution of Integrated Spatial and Energy

d
Planning”. Energies 12, 35.
Destatis (2021). Data on energy price trends. url: https://www.destatis.de/EN/Themes/

we
Economy / Prices / Publications / Downloads - Energy - Price - Trends / energy - price -
trends-pdf-5619002.pdf?__blob=publicationFile (visited on 05/27/2021).
Energiebilanzen, Arbeitsgemeinschaft (2021). Energieverbrauch in Deutschland. url: https:

ie
//www.ag-energiebilanzen.de/20-0-Berichte.html (visited on 06/18/2021).
Garbesi, Karina, Vagelis Vossos, and Hongxia Shen (2012). “Catalog of DC appliances and

ev
power systems”. Lawrence Berkeley National Laboratory, LBNL-5364E.
Gerber, Daniel L, Vagelis Vossos, Wei Feng, Chris Marnay, Bruce Nordman, and Richard
Brown (2018). “A simulation-based efficiency comparison of AC and DC power distribution

r
networks in commercial buildings”. Applied Energy 210, pp. 1167–1187.
Glasgo, Brock, Inês Lima Azevedo, and Chris Hendrickson (2016). “How much electricity can we
er
save by using direct current circuits in homes? Understanding the potential for electricity
savings and assessing feasibility of a transition towards DC powered buildings”. Applied
pe
Energy 180, pp. 66–75.
Johnson, Karl, Vagelis Vossos, Margarita Kloss, Gerald Robinson, and Rich Brown (2016).
Direct Current as an Integrating Platform for ZNE Buildings with EVs and Storage: DC
Direct Systems–A Bridge to a Low Carbon Future? Tech. rep. Lawrence Berkeley National
ot

Lab.(LBNL), Berkeley, CA (United States).

Klingler, Anna-Lena (2017). “Self-consumption with PV+ Battery systems: A market diffusion
tn

model considering individual consumer behaviour and preferences”. Applied Energy 205,
pp. 1560–1570.
Kockel, Christina, Lars Nolting, Rafael Goldbeck, Christina Wulf, Rik W De Doncker, and
rin

Aaron Praktiknjo (2020). “Reducing Environmental Impacts through a Smart Design of

Microgrids: Life Cycle Assessment for AC and DC”. DEStech Transactions on Environment,
Energy and Earth Sciences ICEEE.
ep

Liebowitz, Stan J and Stephen E Margolis (1995). “Path dependence, lock-in, and history”.
Journal of Law, Economics, & Organization, pp. 205–226.
Pr

26

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 Lüth, Alexandra, Jan Martin Zepter, Pedro Crespo del Granado, and Ruud Egging (2018).
“Local electricity market designs for peer-to-peer trading: The role of battery flexibility”.

d
Applied Energy 229, pp. 1233–1243.
Mackay, Laurens, Nils H van der Blij, Laura Ramirez-Elizondo, and Pavol Bauer (2017).

we
“Toward the universal DC distribution system”. Electric Power Components and Systems
45(10), pp. 1032–1042.
Marchionini, Brian and Shuyu Zheng (2018). “Direct Current in Buildings—A look at current

ie
and future trends”. National Electrical Manufacturers Association, Rosslyn, VA, USA.
Noritake, Masatoshi, Kazufumi Yuasa, Takashi Takeda, Hidekazu Hoshi, and Keiichi Hirose

ev
(2014). “Demonstrative research on DC microgrids for office buildings”. 2014 IEEE 36th
International Telecommunications Energy Conference (INTELEC). IEEE, pp. 1–5.
NRW, Energie Agentur (2021). Information Direktlieferung. url: https://www.energieagen

r
tur.nrw/finanzierung/stromvermarktung/direktlieferung?mm=berblick#ts (visited
on 05/11/2021).
er
Pfenninger, Stefan and Iain Staffell (2016). “Long-term patterns of European PV output using
30 years of validated hourly reanalysis and satellite data”. Energy 114, pp. 1251–1265.
pe
Pflugradt, N. (2017). “LoadProfileGenerator”. available at: www.loadprofilegenerator.de.
Rasheed, Amber, Sidra Khan, Hasan Erteza Gelani, and Faizan Dastgeer (2019). “AC vs. DC
Home: An Efficiency Comparison”. 2019 International Symposium on Recent Advances in
Electrical Engineering (RAEE). Vol. 4. IEEE, pp. 1–6.
ot

Sannino, Ambra, Giovanna Postiglione, and Math HJ Bollen (2002). “Feasibility of a DC net-
work for commercial facilities”. Conference Record of the 2002 IEEE Industry Applications
tn

Conference. 37th IAS Annual Meeting (Cat. No. 02CH37344). Vol. 3. IEEE, pp. 1710–1717.
Schlesinger, Michael, Dietmar Lindenberger, and Christian Lutz (2014). “Entwicklung der
Energiemärkte–Energiereferenzprognose”. Projekt Nr. 57/12. Studie im Auftrag des Bun-
rin

desministeriums für Wirtschaft und Technologie. Basel/Köln/Osnabrück: ewi/gws/prognos.

Ullah, Shahid, Ahmed MA Haidar, and Hushairi Zen (2020). “Assessment of technical and finan-
cial benefits of AC and DC microgrids based on solar photovoltaic”. Electrical Engineering,
ep

pp. 1–14.
Vossos, Vagelis, Karina Garbesi, and Hongxia Shen (2014). “Energy savings from direct-DC in
US residential buildings”. Energy and Buildings 68, pp. 223–231.
Pr

27

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039
 Vossos, Vagelis, Karl Johnson, Margarita Kloss, Mukesh Khattar, Daniel Gerber, and Rich
Brown (2017). “Review of dc power distribution in buildings: A technology and market

d
assessment”. LBNL Report, Berkeley, CA, USA.
Weiss, Roland, Leopold Ott, and Ulrich Boeke (2015). “Energy efficient low-voltage DC-grids

we
for commercial buildings”. 2015 IEEE First International Conference on DC Microgrids
(ICDCM). IEEE, pp. 154–158.
Weniger, Johannes, Tjarko Tjaden, and Volker Quaschning (2014). “Sizing of residential PV

ie
battery systems”. Energy Procedia 46, pp. 78–87.
Zepter, Jan Martin, Alexandra Lüth, Pedro Crespo Del Granado, and Ruud Egging (2019).

ev
“Prosumer integration in wholesale electricity markets: Synergies of peer-to-peer trade and
residential storage”. Energy and Buildings 184, pp. 163–176.

r
er
pe
ot
tn
rin
ep
Pr

28

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4083039