Optimal household energy management and economic analysis: from sizing to operation scheduling

Full text

Page 1

LABORATOIRE D’ECONOMIE DE LA PRODUCTION
ET DE L’INTEGRATION INTERNATIONALE

UMR 5252 CNRS - UPMF

______________________


LEPII
BP 47 - 38040 Grenoble CEDEX 9 - France
1221 rue des Résidences - 38400 Saint Martin d'Hères
Tél.: + 33 (0)4 76 82 56 92 - Télécopie : + 33 (0)4 56 52 85 71
lepii@upmf-grenoble.fr - http://www.upmf-grenoble.fr/lepii



_____________________




CAHIER DE RECHERCHE

N° 11



Optimal household energy
management and economic analysis :
from sizing to operation scheduling



T.T. Ha Pham1
C. Clastres1,2
F. Wurtz1
S. Bacha1
E. Zamaï3


juillet 2008


1 Grenoble Electrical Engineering (G2ELAB) - INPG, UJF, CNRS
2 LEPII – UPMF, CNRS
3 Laboratoire des sciences pour la conception, l'optimisation et la production (G-SCOP) - INPG, UJF, CNRS

halshs-00323581, version 1 - 22 Sep 2008
Author manuscript, published in "Advances and Applications in Mechanical Engineering and Technology 1, 1 (2010) 35-68"

Page 2

halshs-00323581, version 1 - 22 Sep 2008

Page 3

1
Optimal household energy management and economic
analysis: from sizing to operation scheduling

T.T. Ha Pham1, C. Clastres1,2,*, F. Wurtz1, S. Bacha1, E. Zamaï3

1 Grenoble Electrical Engineering (G2ELAB) CNRS - INPG / UJF
BP 46-38402 Saint Martin d'Heres, France

2 LEPII-EREN - University of Grenoble II
BP 47 - 38040 Grenoble Cedex 9 - France

3 Laboratoire des sciences pour la conception, l'optimisation et la production (G-SCOP) CNRS - INPG / UJF
46, avenue Félix Viallet - 38031 Grenoble, France


July 2008

Abstract
The study presented in this paper takes part in a project aiming to increase the value of solar production for
residential application with a medium-term vision where preferential solar energy subsidies will decrease
before to disappear. This study is dedicated to propose and develop optimal energy architecture at supply
side, a multi-source system based on photovoltaic (PV) solar energy connecting to main electrical network,
taking further into account the effectiveness of intelligent demand side management. To investigate this issue,
a method of optimal supplying system sizing and household energy management has been developed. This
method, which has been formulated employing Mix Integer Linear Programming (MILP), enables the
calculation of the appropriate configuration for power supply system and the optimal operation control to be
applied. Using a Net Present Value (NPV) and Probability Index (P.I) basis, the economic analysis allows
estimation of the viability of the proposed system under different factors of influence such as renewable
energy policies, technology evolutions leading to cheaper installed PV module cost and deregulated
electricity market. Simulation results show that, the solution makes it possible for PV power to be
significantly valued by the customers without subsidized measures.
Keywords: connected-grid PV system, battery storage, sizing optimization, energy management, MILP

* Corresponding author
Tel. : +33 (0)476826430; Fax : +33 (0)476826300
E-mail address : cedric.clastres@g2elab.inpg.fr
This work was supported by the French National Research Agency - PV Program and is included in the MULTISOL
Project.
halshs-00323581, version 1 - 22 Sep 2008

Page 4

2
Nomenclature

Model parameters
ηB
CB
cg(t)
cgn
Cond
CPV
cs(t)
i
k
rch
rdch
SP(t)
t
T
Battery storage efficiency
Coefficient cost of storage system, [€/kWh]
Grid energy price, [€/kW]
Coefficient of grid connection fix cost, [€/kW]
Coefficient cost of inverter, [€/kW]
Coefficient cost of peak power rate of PV module, [€/kWp]
Sellback price of local production, [€/kWh]
Discount rate, [%/year]
Study year, [year]
Charge coefficient rate, [kW/h]
Discharge coefficient rate, [kW/h]
Spot price, [€/kWh]
Calculation step time, [hour]
Study period, equal to lifetime of installation, [year]
Decision variables
α(t) Decision binary variable, α(t) = 1 if the battery is in the charging mode, α(t) = 0 if the battery
in the discharging mode
β(t) Decision binary variable, β(t) = 1 if the system imports grid energy, β(t) = 0 if the system
purchases its PV production
ω(t)
Decision variable, used to translate the absolute relation into linear representation
Pbin(t) Charge power consign [kW]
Pbout(t) Disharge power consign [kW]
Pg(t) Consumed grid power [kW]
PL(t) Electrical demand prevision [kW]
PLP(t) Consumed power by controllable loads
PNLP(t) Consumed power by non-controllable loads
PPV(t) PV available power [kW]
Ps(t) PV power to be used locally [kW]
SOC(t) Temporal state of charge [kWh]
z(t) PV power to be injected to network [kW]


1. Introduction

In Europe, the rapid growth connected-grid PV systems in the residential sector over the last year have
been promoted by government supported programs with considerable investment subsidies [1]. By this way,
the PV power producers find it currently beneficial to sellback maximum of his solar production to electricity
utilities. This is especially convenient for short-time frame when PV penetration rate in electrical network still
halshs-00323581, version 1 - 22 Sep 2008

Page 5

3
remains marginal or to be applied in the small scale applications.

However, actual policies to develop the use of renewable energies would evolve. Feed-in tariffs could be
lower to lessen the cost of renewable energies promotion. They would be then disappeared (as in Japan, for
example) or replaced by another system less attractive (green certificates or market component as in Spain
[2]). Given that context, it will be difficult to justify the real interest of this operation mode of PV connected-
grid system as currently.

However, the potential benefits of PV system are quite large. The core idea is that PV applications could
impact both supply-side and demand-side issues. Aside being an alternative for the housing power supply
having lower environment impacts, PV also serve as a vehicle for triggering efficient energy utilization by
influencing consumer awareness of energy saving.

Another observation and analysis of energy markets changing also make connected-grid PV system
interesting to study. The liberalization process leads to the end of regulated tariffs (toward 2011 in France).
So, market price contracts would be generalized to all consumers, with certainly an indexation on volatile
spot prices. A consumer could have incentives and benefits to develop PV generation to limit the price risk
volatility. This is, for a household consumer, a mean to hedge it against high prices.

Lastly, energy vulnerability of several European countries and climate change policies imply to develop
cleaner production and demand-side management (rationale use of energy, energy positive houses as in Great-
Britain where all new houses in 2016 would be zero Greenhouse Gas emission [3]).

These issues at stake ask for investigating the innovative energy architecture for housing applications. It is
expected that the building of tomorrow should be a positive, green and intelligent element. The project
sustaining this paper, MULTISOL, has the ambition to design an economical and technical efficient
framework for household energy management. A new architecture is proposed with: at supply side, a PV-
based multi-source system; and at demand side, source and load co-management.

2. Model framework description

In this project, the works to design and demonstrate the multi-sources and loads co-management system
architecture has been in progress. We suggest separating the components between the house and electric
network into two connection boards (Fig. 1), [4]. All production means such as PV generator, battery,
network, and possible diesel or others complementary sources, are connected in the "Power production
control board" to supply loads via the traditional electric delivery box called "Power delivery control board".
A coupling and

∗ STC is specified at condition air mass of 1.5, radiation of 1 kW/m2, and ambient temperature of 25°C.
halshs-00323581, version 1 - 22 Sep 2008