Dimensioning of electricity storage according to small wind turbine power generation and household load patterns

Abstract

This paper analyzes wind speed data measured in the Tallinn-Harku Aerological Station (Estonia). Dimensioning of a wind turbine and electricity storage for a typical household is discussed. The final part describes electricity storage dimensioning based on a combination of Nord Pool Spot (NPS) prices and grid connected household wind turbine generation.

Full text

Dimensioning of Electricity Storage according to
Small Wind Turbine Power Generation and
Household Load Patterns
Argo Rosin, Ivo Palu, Kai Rosin, Aivar Auväärt
Tallinn University of Technology
argo.rosin@ttu.ee, ivo.palu@ttu.ee, kairosin76@gmail.com, aivar.auvaart@ttu.ee
Abstract- This paper analyzes wind speed data measured in the
Tallinn-Harku Aerological Station (Estonia). Dimensioning of a
wind turbine and electricity storage for a typical household is
discussed. The final part describes electricity storage
dimensioning based on a combination of Nord Pool Spot (NPS)
prices and grid connected household w ind turbine generation.
I. INTRODUCTION
The use of large wind turbines is common almost in every
European country, at the same time expansion of small wind
turbines expends great efforts. There are many small wind
turbine manufacturers and importers, but due to the relatively
high cost of the small-scale devices and lack of financial
support from the governments the market is still
inconsiderable.
Both the payback time of the investment and the technical
practicability to use the solution are to be taken into account.
Small wind turbines discussed in this article are considered to
be in the range of 1 – 20 kW, with typical tower height not
exceeding 10-20 meters. Wind speeds closer to ground are
always smaller than those at the heights of MW-scale wind
turbines since they are more affected by surrounding
obstacles [1].
In order to comply with power peaks, to reduce the
installed generation capacity and to balance missing long- and
short-term coincidence between power generation and
demand, energy storage (ES) is a crucial component of
distributed generation systems [2]. Optimal dimensioning of
electricity storage according to the energy production of
micro-scale renewables (in residential areas and households)
and electricity consumption are important topics in the
development of micro- and smartGRID technologies to
increase system reliability and to reduce the profitability time.
Opening of electricity market gives new opportunities to
renewable energy sources. Optimization of renewable
systems (including hybrid renewable systems and wind
turbines with energy storages) have been studied in detail in
Estonia [3-6] and in world [7], however, only few analyses
are available about small-scale wind energy systems
(including feasibility analysis, dimensioning of energy
storages and small-scale wind turbines) for households, and
optimization of energy storages of wind turbine systems
according to open electricity market prices (e.g. Nord Pool
Spot).
This article reports on the feasibility of small wind turbines
to cover household electricity consumption. No direct
attention is paid to the payback time or cost of the investment.
Main value of this article is to present the possible size of the
wind turbine and possible storage capacity for load coverage
of an average household (also in open-market conditions).
II. WIND SPEED ANALYSIS
In Estonia (also in Scandinavia) the strongest average wind
is in autumn and winter months, especially during November,
December and January, when the average wind speed is about
8 m/s on coastal areas and 5 m/s on inland. The following
analysis shows us similar seasonal differences. The analysis
below is based on the wind speed data measured in the
Tallinn-Harku Aerological Station (latitude N 59°23´53´´;
longitude E 24°36´10´´, height above sea level 33 m), and on
average household energy consumption data described in [8].
As a result of the wind speed analysis, the average wind
speed of an average day in July compared to January is up to
1.46 times lower. In July at the peak hour (12 o’clock), the
average wind speed is up to 1.12 times lower than in
December at the peak hour (11 o’clock). In July at the lowest
hour (21 o’clock), the average wind speed is up to 2.16 times
lower than in December at the peak hour (23 o’clock). About
61% of the resource is concentrated on the winter season
from October until March, when the average hourly wind
speed is mostly over annual average (Fig. 1).
1,5
2,0
2,5
3,0
3,5
4,0
4,5
00:00
02:00
04:00
06:00
08:00
10:00
12:00
14:00
16:00
18:00
20:00
22:00
v (wind speed), m/s
t (time), hours
Jan Feb Mar Apr May
Jun Jul Aug Sep Oct
Nov Dec Averag e
Fig. 1. Average daily wind speed by months (Harku 2005-2009)

As shown in Figure 2, during the last five years, about 27%
of hours the average wind speed was above 3 m/s.
For further analysis the wind turbine with the rated power
of 5 kW was chosen. The power characteristic of the wind
turbine can be presented as the fourth order polynomial
function (Fig. 3) of wind speed (R-squared value of the trend
is 0.997). As a result of the energy generation analysis, the
average energy generation of an average day in July
compared to January is up to 2.8 times lower.
1,3%
9,1%
27,4%
50,1%
71,1%
85,9%
94,2%
97,9%
99,3%
99,7%
99,9%
100,0%
100,0%
100,0%
0%
20%
40%
60%
80%
100%
120%
140%
0
2000
4000
6000
8000
10000
12000
0
1
2
3
4
5
6
7
8
9
10
11
12
More
Frequency
v (wind speed), m/s
Frequency Cumulative %
Fig. 2. Histogram of wind speed (Harku 2005-2009)
y = 0,4396x
4
- 25,43x
3
+ 453,71x
2
- 2295,4x + 3765,6
R
2
= 0,997
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
v (w ind speed), m/s
P (power), W
P=f ( v ) Poly. (P=f(v))
Fig. 3. P=f(v) characteristic of the chosen wind turbine
III. WIND TURBINE DIMENSIONING FOR TYPICAL
HOUSEHOLDS
The following calculations (1-2) are simplified and do not
take into account the wind-turbine performance ratio,
including system losses and efficiency of power electronics.
3
32
2
1
8vAc
vd
Pp⋅⋅⋅⋅=
⋅⋅⋅⋅
=
ρ
ρηπ
, (1)
g
c
grcr E
E
PP ⋅= ,, , (2)
where A – rotor area;
ρ
- density of air;
η
- efficiency;
d – diameter of rotor;
v
- wind speed; cp – power coefficient;
Ec – electricity consumption per day; E
g – electricity
generation of a wind-turbine per day; Pr,g – rated power of a
wind turbine; Pr, c – needed rated power of a wind turbine for
covering daily electricity consumption.
Average electricity consumption (in common household of
Estonia) is on average day (about 0.5 kWh per hour) [9],
without consumption of an electrical water heater, in July can
be covered by a wind turbine (with the described P=f(v)
characteristic) with the rated power of 12.56 kW (Fig. 4). To
cover average electricity consumption in January, a wind
turbine with the rated power of 4.46 kW should be installed.
In accordance with the average holiday/weekend (HD) and
workday (WD), electricity consumption (0.66 kWh/h and
0.38 kWh/h) of the rated power of wind turbines was
calculated for an average day of each month. It was found
that the smallest electricity generation is in July, and the
highest one is in January. Based on the calculations, the
highest rated power of the chosen wind turbine (in
accordance with the wind turbine characteristic) should be
approximately 17 kW (holiday in July) and the lowest one 3.4
kW (workday in January) (Fig. 4).
4,5
6,7
8,2
7,5
8,5
8,6
12,6
10,7
8,4
6,3
5,1
4,8
3,4
5,1
6,3
5,7
6,4
6,5
9,6
8,1
6,4
4,8
3,9
3,7
5,9
8,9
10,8
9,8
11,1
11,3
16,6
14,1
11,1
8,4
6,7
6,4
0
2
4
6
8
10
12
14
16
18
20
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Rated power, kW
Month
Average day (A D) Workday (W D) Week end/holi day (HD)
F
ig. 4. Rated power of a wind turbine according to the monthly wind speed
Based on annual average electricity generation and
household consumption, the rated power of a wind turbine
should be approximately from 6.5...7.6 kW. The annual
average rated power of a wind turbine should be 5.84 kW for
a workday (WD) and 10.08 kW for a weekend (HD). On
workdays the energy consumption is much lower than on
weekends. The rated power of a wind turbine was chosen in
accordance with the annual average rated power of a
weekend, which is approximately 10 kW. To have rated
power of 10 kW, two wind turbines with the characteristic
described in Fig. 3 should be used. As shown, the rated power
on workdays in each month is lower than 10kW. This means
that theoretically it is possible to store surplus of energy on
workdays and to use it on weekends. Practically it does not
work in this way, while there are long zero-generation
periods, which should be covered from an additional energy
source. Our further calculations are based on 2 x 5 kW wind
turbines. Two wind turbines with the rated power of 5 kW
will have theoretically the maximum total day generation of
27 kWh (January) and the minimum 9.6 kWh (July).
In different seasons the deviation of energy generation of
wind turbines compared with a PV-system is relatively low

[9]. The coefficient of variation VR of monthly generation is
25% (for a PV system 72%) (3).
wt
n
i
wtiwt
REn
EE
V
⋅
−
=
∑
=1
,, (3)
where Ewt,i – generated electricity at the hour i; n – 24 hours
in a day;
E
wt – average daily electricity generation.
As compared to a PV system, lower variations in annual
electricity generation are the main advantage to use wind
turbines.
In an OFF-grid system for load coverage in the winter
season, when there exist long zero-generation periods (up to
150 hours), it is reasonable to use additionally a micro-CHP
or a diesel generator. A micro-CHP produces additional
thermal energy, which can be fully used in the winter season.
In an ON-grid system, in the winter season covering the
shortage of electricity with low-tariff energy stored in the
Wind-system energy storage is a suitable solution. Next,
energy balance of a household wind energy system and
energy reserve dimensioning for a storage system are
analyzed.
IV. ELECTRICITY RESERVE DIMENSIONING OF A HOUSEHOLD
WIND ENERGY SYSTEM FOR LOAD COVERAGE
Energy balance of a wind energy system can be described
according to the following simplified formula (4):
lossp
E
resdirwt
losspcwt
EEEEE
EEEE
c
+++=
++=
, (4)
where Ewt – electricity generated by a wind energy system;
Ec – electricity consumption; Esp – surplus of generated
electricity; Elos – total losses; Edir – direct consumption of
electricity generated by a wind system; Eres – indirect
consumption of electricity generated by a wind energy system
(stored energy reserve of wind turbine generated energy).
In the calculations system losses are not taken into account
(Elos = 0).
A. Balance between generation and load
The first step to define the capacitance needed for
electricity reserve for load coverage of a household wind-
system is the analysis of balance between wind turbine
generation and load consumption on an average day of each
month (5). While the WD and HD have different
consumption curves, the analysis should be made separately
for both days.
iwticibal EEE ,,, −= , (5)
where Ebal,i – energy balance at the hour i.
According to the calculations (5) at WD, the surplus of
generated energy is the highest on midday, when the load is
trivial. Load maximum prevails in the evening from 16 to 21
o’clock. This means that generated energy should be stored
for the evening period (Fig. 5).
Measures should be taken to solve this surplus problem. On
HD the direct load coverage is better than on WD. In the
summer season main problems are shortage and higher needs
for energy reserves as compared to the winter season. In the
summer season energy reserves needed on HD are similar to
those needed on WD. (Fig. 6).
-1,5
-1,0
-0,5
0,0
0,5
1,0
1,5
00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
hours
kWh
Jan Feb Mar Apr
May Jun Jul Aug
Sep Oct Nov De c
WD average
Fig. 5. Balance of WD consumption and generation
-1,5
-1,0
-0,5
0,0
0,5
1,0
1,5
00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
hours
kWh
Jan Feb Mar Apr
May Jun Jul Aug
Sep Oct Nov De c
HD average
Fig. 6. Balance of HD consumption and generation
B. Electricity surplus and shortage
The analysis (6) below shows that a wind turbine with the
rated power of 10 kW can cover WD electricity consumption
from January to December (Fig. 7). On HD electricity
consumption can be covered from October to February.
While in the calculations used electricity data is measured in
February and March, and electricity consumption in the
summer season is reduced about 20...30%, it can be
theoretically expected that a wind turbine with the rated
power of 10kW can cover also consumption of the summer
season.
24;
)(
1
,
,
1
,
=
−
==
∑
∑
=
=n
E
EE
E
E
kn
i
ic
ic
n
i
iwt
c
sp
sp , (6)

where Ewt,i – generated electricity at the hour i; n – 24 hours
in a day; Ec,i – electricity consumption at the hour i.
Without losses the average annual surplus of electricity
generation of a wind turbine-system on holidays and
workdays is 8.5% and 87.5%, accordingly. In the summer
season an average surplus is 10.3%, on WD and HD,
accordingly 44.5% (surplus) and -16.4% (shortage). In the
winter season the average surplus is 76%, on WD and HD,
accordingly 130.5% and 33.4%.
124%
48%
22%
34%
18%
17%
-20%
-6%
19%
58%
97%
107%
194%
94%
60%
75%
55%
53%
4%
23%
56%
107%
158%
171%
70%
12%
-8%
2%
-10%
-11%
-40%
-29%
-10%
20%
49%
57%
-100%
-50%
0%
50%
100%
150%
200%
250%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
AD WD HD
Fig. 7. Surplus/shortage of an average day of a month
C. Dimensioning of electricity reserve in accordance with average
energy generation
Approximately 39% of wind turbine generated energy can
be directly used on workdays and 72% on holidays. This is
about 73% of a workday and 78% of holiday total electricity
consumption (7-9).

ires
ic
E
iwt
E
E
iwticicdir EEEE
,
,,
)( ,,, ∑∑
≤
−−=
, (7)
∑∑
>≤
+=
iciwticiwt EE
ic
EE
iwtdir EEE
,,,,
,, , (8)
where Edir – directly from wind turbine consumed
electricity.
c
dir
dir E
E
k=, (9)
About 56% of annual average wind turbine generated
energy is used directly, which makes approximately 80 % of
the annual average consumption.
The easiest way to calculate the energy reserve needed
(storage capacitance) for indirect load coverage is based on
the difference of average hourly electricity generation and
consumption (10 - 11).
Depending on the consumption pattern, about 15 to 20 %
of the generated energy should be stored in energy storage,
making up 22 to 27 % of the consumption.
⇓
≤,
,, iciwt EE , (10)
∑∑∑
===
−=−==
n
i
iciwtiwtic
n
i
n
i
iresres EEEEEE
1
,,,,
11
,)( , (11)
where Eres,i – energy reserve needed at the hour i; Eres –
average daily energy reserve needed.
On WD, the highest energy reserve is needed in July (5.29
kWh) and the lowest in January (0.57 kWh) (Fig. 8). On HD,
in turn, the highest energy reserve needed is 7.96 kWh and
the lowest is 1.19 kWh (Fig. 9).
17,9
8,7 5,5 7,0 5,1 4, 9
0,4 2,1
5,2
9,9
14,6 15,8
8,7
7,3
6,7 5,5
5,0 4,6
3,9
4,1
5,6
7,3
8,4 8,3
0,6
1,9
2,5
3,8
4,3 4,6
5,3
5,2
3,6
1,9
0,8 0,9
-5
0
5
10
15
20
25
30
Jan Feb Mar
A
pr May Jun Jul
A
ug Sep O c
t
Nov Dec
kWh
Mon th
Surplus/shortage Di rect c onsumption Ind irect consumption
Fig. 8. Direct and indirect load coverage of an average WD of a month
11,2
2,0 -1,2 0,2 -1,6 -1,8 -6,3 -4,6 -1,5
3,1
7,9 9,0
14,8
12,1
12,0 11,8 11,1 10,8 8,0 9, 0 11,3
12,7
14,2 13,7
1,2
3,9
4,0 4,1 4,8 5,1 8,0 7,0 4, 7
3,3
1,8 2,3
-10
-5
0
5
10
15
20
25
30
Jan Feb Mar
A
pr May Jun Jul
A
ug Sep Oct No v Dec
kWh
Month
Surp lus /shortage Direct consumption Ind ir ect consumption
Fig. 9. Direct and indirect load coverage of an average HD of a month
D. Dimensioning of electricity reserve in accordance with
duration of periods without power generation
Another calculation method of the energy reserve, however
rarely used, is based on the analysis of the frequency of the
duration of hours without power generation (PG) and average
electricity consumption. The following histogram (Fig. 10)
shows that 92% of periods without power generation are
shorter than 24 hours. In accordance with Figure 10, over
90% of periods without power generation are shorter than 22

hours. In accordance with the presented histogram, the energy
reserves for an average day, workday and holiday should be
11.1, 8.5 and 15 kWh. It is almost two times more than
described above.
Another possibility to define an energy storage needed for
load scheduling is to take into account the longest daily
average periods without power generation. The longest
average periods without power generation (Fig. 11) per day
are in July (11.15 hours) and in January (10.62 hours). Based
on an average daily (0.504kWh/h), workday (0.385 kWh/h)
and holiday (0.665kWh/h) consumption (without electrical
water heater) [9] and the longest monthly average period
without power generation, the calculated energy reserves for
a period without power generation in July on an average day
should be accordingly 5.6, 4.3 and 7.5 kWh. Calculated
energy reserves are quite similar to the energy reserves
described in Figures 8 and 9.
35%
48%
56%
61%
66%
71%
76%
81%
86%
89%
91%
92%
100%
0
100
200
300
400
500
600
700
800
2
4
6
8
10
12
14
16
18
20
22
24
More
Frequency
Bin
Frequency Cumulative %
Fig. 10. Histogram of hours without power generation (2005-2009)
It can be seen from Figure 10 that over 7 % of hours
without energy generation are longer than 24 hours. It is
useful to have an overview about these periods to plan an
additional energy source. The longest periods without power
generation calculated from the measured wind speed (from
2005 to 2009) in accordance with the characteristics of the
wind turbine were in January 2006 (146 hours) and in
October 2007 (155 hours). The longest shortage in power
generation could be over 6 days (Fig. 11). According to the
longest shortage and possible weighted average consumption
(including consumption of workdays and holidays), the
additional energy source should cover during these days
energy consumption up to 80 kWh and peak power up to 8
kW. With priority based load control the needed peak power
could be reduced up to 4 times (real reduction depends on the
used appliances and customer demands).
For a customer another interesting value is annual working
hours of the additional energy source to cover the shortage of
wind energy. In accordance with average maximum shortages
of each month (Fig. 12), the total annual need for an
additional energy source could be at least 670 h
(approximately equal to 350 kWh of average electricity
consumption).
According to the calculations, the highest energy
generation was found in January and the lowest in July. As
seen in Figure 13, the average maximum hours without PG
are the longest in January and the shortest in July. The
relation between average maximum hours without PG to the
total hours in a month shows that probability of longer
periods without PG is higher than in other months. This
phenomenon shows us that better PG in January does not
mean lower demands for an additional power generation
system.
10,62
8,78
11,15
8,64
146
45
155
0
20
40
60
80
100
120
140
160
180
8
9
9
10
10
11
11
12
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Maxi mum hours w ithout PG
Average hours wit hout PG
Month
Average hours without PG Max imum hours without PG
Fi
g. 11. Average and maximum hours without power generation (2005-2009)
30,3%
16,0%
19,3%
10,9%
10,7%
9,1%
9,0%
12,7%
20,6%
21,2%
21,4%
23,4%
74,6
51,6
70
35,4
42,8
35,4
41,2
51,8
75,6
71,4
54
63,4
0
10
20
30
40
50
60
70
80
90
100
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Average of maximum hours without
PG
tmax/t
Month
tmax/t Average of maximum hours without PG
Fig. 12. Average maximum hours without PG and relation of average
maximum hours to average total hours without PG (2005-2009).
V. ELECTRICITY RESERVE DIMENSIONING FOR ON-GRID
CONNECTED WIND TURBINE SYSTEM BY COVERING LULLS WITH
LOW SPOT PRICE
Average NPS (Nord Pool Spot) price during the measured
period (April 2010 to March 2011) in the EE (Estonia) area
was calculated as 46.30€/MWh. The NPS price curve is
similar on workdays and at weekends. The average price of
workdays is 47.81 €/MWh and that of weekends is
42.62€/MWh. The main difference of WD and HD curves is
higher midday peak on WD and higher evening peak on HD).
More detailed analysis of the first half year is described in
[10], and profitability of energy storages is described in [11].
The analysis below should show if it is possible to use a
calculated energy reserve to cover lulls with cheaper
electricity from spot market. The following electricity reserve
calculations are based on hourly prices of an average day in
months. If the hourly price is lower than the average WD or
HD price of the month, the electricity consumption and

storage charging are covered from the grid. If the hourly WD
or HD price is higher than the average price of the month, the
electricity consumption is covered by electricity storage (12).
∑
=
=⇒>
n
i
icresmim EEpp
1
,, , (12)
where pm,i – average price of the hour i in a month m, m
p –
average price of a month m
Based on the described calculations (14), the highest
needed energy reserve is required on holidays from May to
August, which is 14.4 kWh (Fig. 13). On workdays the
highest energy reserve, 7.7 kWh, is needed in May.
5,3
6,7
5,4
4,2
7,7
6,5
1,7
0,8
6,5
6,7
5,2
5,3
12,7
10,7
12,8
13,2
14,4
14,4
14,4
14,4
13,6
8,9
12,8
12,9
0
2
4
6
8
10
12
14
16
18
20
Jan Feb Mar Ap r May Jun J ul Aug Sep Oc t No v Dec
Capac ity o f energ y reserve, k Wh
Mo nths
WD HD
Fig. 13. Energy reserve based on monthly average WD and HD prices
If we compare an electricity reserve in accordance with the
duration of periods without power generation (Fig. 10), it can
be concluded that for wind turbine calculated electricity
reserve (15 kWh) it is sufficient to schedule the load
according to the average spot price.
VI. CONCLUSION
According to average electricity consumption and
depending on the monthly average wind speed, the rated
power of a wind turbine should be chosen from 5 to 20 kW.
Over 90% of periods without power generation are shorter
than 22 hours. Energy reserves for an average day, workday
and holiday should be 11.1, 8.5 and 15 kWh. It is also
sufficient to schedule the load according to average spot
prices.
Different methods used for electricity storage dimensioning
give completely different and inappropriate results. For
example, calculations based on periods without power
generation give us more realistic results than the results
calculated according to the average hourly wind speed and
consumption at an average day of a month. The longest
periods without power generation calculated from the
measured wind speed (from 2005 to 2009) in accordance with
the characteristics of the wind turbine were in January 2006
(146 hours) and in October 2007 (155 hours). In accordance
with average maximum shortages of each month, the total
annual need for an additional energy source could be at least
670 h (approximately equal to 350 kWh of average electricity
consumption).
The unbalance between consumption and power generation
has the highest impact on the dimensioning of storage
capacitance. Profitability of a wind turbine depends mostly on
the price of electricity and the consumption pattern. To assure
the shortest profitability time, electricity consumption and
real-time dynamic price should be increased and decreased
synchronously with the wind turbine generation.
Today in northern regions, such as Baltic countries, small
wind turbines are only feasible in OFF-grid systems, where
the grid connection is not economically feasible and an
additional energy source (e.g. diesel generator) for balancing
is available. In ON-grid wind turbine systems, it is more
feasible to cover shortage with cheaper energy stored from
the grid in the OFF-peak time. For Nordic countries,
additional investigations are needed to find out if it is more
profitable to combine PV power with wind power or if it is
more feasible to use microCHPs combined with wind or with
PV power. Also, the intelligent real-time price based load
scheduling possibilities of households should be taken into
account.
ACKNOWLEDGMENT
Authors thank the Estonian Ministry of Education and
Research (Project SF0140016s11) and Estonian Archimedes
Foundation (Project “Doctoral School of Energy and
Geotechnology II“) for financial support to this study.
REFERENCES
[1] Manwell, J. F., McGowan, J. G., Rogers, A. L., Wind Energy
Explained: Theory, Design and Application, Second Edition, Wiley,
2009, pp 689.
[2] Matics, J.; Krost, G.; , "Intelligent Design of PV based Home Supply
using a Versatile Simulation Tool," Intelligent Systems Application to
Power Systems, 2005. Proceedings of the 13th International Conference
on , vol., no., pp.61-66, 6-10 Nov. 2005
[3] Põder, Vahur; Lepa, Jaan; Palge, Veli; Peets, Tõnis; Annuk, Andres
(2009). The Estimation of Needed Capacity of a Storage System
According to Load and Wind Parameters. Oil Shale, 26(3), 283 - 293.
[4] Palu, I.; Oidram, R.; Keel, M.; Tammoja, H. (2009). Balancing of wind
energy using oil-shale based power plants at erroneous wind forecast
conditions. Oil Shale, 26(3 S), 189 - 199.
[5] Agabus, H.; Tammoja, H. (2009). Estimation of wind power production
through short-term forecast. Oil Shale, 26(3 S), 208 - 219.
[6] Pertmann, I. (2011). Estonian Wind Farms Need for Full Balance
Power. Oil Shale, 28(1S), 193 - 202.
[7] Borowy, B.S.; Salameh, Z.M.; , "Methodology for optimally sizing the
combination of a battery bank and PV array in a wind/PV hybrid
system," Energy Conversion, IEEE Transactions on , vol.11, no.2,
pp.367-375, Jun 1996
[8] Rosin, A.; Hõimoja, H.; Möller, T.; Lehtla, M.; , "Residential electricity
consumption and loads pattern analysis," Electric Power Quality and
Supply Reliability Conference (PQ), 2010 , vol., no., pp.111-116, 16-18
June 2010.
[9] Rosi n, A.; Rosin, K.; Auvaart, A.; Strze lecki , R. "Di mensioning of
household electricity storage for PV-systems and load scheduling based
on Nord Pool Spot prices." Przegląd Elektrotechniczny 88.4B
(2012):37-40.
[10] Auväärt, A.; Rosin, A.; Belonogova, N.; Lebedev, D.; , “NordPoolSpot
price pattern analysis for households energy management,” 7th
International Conference-Workshop Compatibility and Power
Electronics, 2011 , pp.103-106, June 01 - 03, 2011.
[11] Rosin, A.; Auväärt, A.; Lebedev, D. (2012). Energy storage
dimensioning and feasibility analysis for household consumption
scheduling based on fluctuations of Nord Pool Spot price. Przeglad
Elektrotechniczny, 88(1a), 37 - 40.