Building of Sweden Yang Zhang , Anders Lundblad, Pietro Elia - - PowerPoint PPT Presentation

Comparative Study of Battery and Hydrogen Storages to Increase Photovoltaic Self-sufficiency in a Residential Building of Sweden

Yang Zhang , Anders Lundblad, Pietro Elia Campana, Jinyue Yan KTH-Royal Institute of Technology April, 20th, 2016

Introduction

• PV capacity in Europe and Sweden
• PV capacity in Italy, Greece and Germany can meet more than 7% of electricity demand[1];
• 0.06% in Sweden[2]

[1] IEA, Snapshot of Global PV Markets. 2015. [2] National Survey Report of PV Power Applications in SWEDEN. 2014.

• Local weather condition;
• Low incentives;
• PV intermittence;
• Daily Mismatch;
• Seasonal Mismatch;
• etc.
• Reasons at end user side

Introduction

• Local weather condition;
• Low incentives;
• PV intermittence;
• Daily Mismatch;
• Seasonal Mismatch;
• etc.
• Reasons at end user side

0,2 0,4 0,6 0,8 1 1,2 Self-Consumed Electricity Surplus Electricity

Electricy Price, SEK/kWh

Elspot Price Grid charge, Energy Tax, etc.

Revenue from PV panels

US$0.13/kWh US$0.03/kWh

Introduction

• Local weather condition;
• Low incentives;
• PV intermittence;
• Daily Mismatch;
• Seasonal Mismatch;
• etc.
• Reasons at end user side

Averaged Daily Consumption in Different Months Averaged Daily Production in Different Months

Introduction

• PV capacity in Europe and Sweden
• PV capacity in Italy, Greece and Germany can meet more than 7% of electricity demand[1];
• 0.06% in Sweden[2]

[1] IEA, Snapshot of Global PV Markets. 2015. [2] National Survey Report of PV Power Applications in SWEDEN. 2014.

• Local weather condition;
• Low incentives;
• PV intermittence;
• Daily Mismatch;
• Seasonal Mismatch;
• Reasons at end user side

Storage Battery Storage Hydrogen Storage

System schematic layout

System schematic layout: Battery Storage System

Load Profile PV model Weather Profile Battery characterizing parameter Operation Strategy PV Size Battery Size PV LCC Battery LCC Simulation Hourly PV Production Battery V-I Model Battery Lifetime Model Saving & Income Net Present Value Electricity Price Profile Self Sufficiency Ratio

System schematic layout: Hydrogen Storage System

Load Profile PV model Weather Profile Operation Strategy PV Size Electrolyzer Size PV LCC Hydrogen System LCC Simulation Hourly PV Production Electrolyzer Saving & Income Net Present Value Self Sufficiency Ratio Fuel cell Tank size Fuel Cell Size Electricity Price Profile

• Cost
• Investment Cost
• Replacement Cost
• Maintenance Cost
• Self Consumption Revenue
• Export Revenue

25 , , 1 1

( ) (1 )

y OM y R y y y r

R C C NPV I d

 

    



Net Present Value PV Size Battery Size Electrolyzer Size Fuel Cell Size Hydrogen Tank Size System Performance Self Sufficiency Ratio Trade Off Optimization Genetic Algorithm

• Revenue

System schematic layout

2 1 2 1

max( ( ),0) 1 100% ( )

t t t t

G t SSR L t             

 

Net Present Value & Self Sufficiency Ratio

25 , , 1 1

( ) (1 )

y OM y R y y y r

R C C NPV Inv d

 

    



2 1 2 1

max( ( ),0) 1 100% ( )

t t t t

G t SSR L t             

 

Component Cost Life Time O&M Reference PV Panels 12900 SEK/kW 25 Years 0.01 [1] Fuel Cell 4000 USD/kW 30000 Working Hours 0.005 [14] Electrolyzer 5000 USD/kW 15 Years 0.005 [14] Hydrogen Tank 570 (USD/kg) 15 Years 0.005 [14] Lithium ion battery 469 USD/kWh 746 @ 80% DOD 0.005 [9] Net Present Value Self Sufficiency Ratio

• Cost
• Revenue

Item Cost Self Consumption Revenue Elspot Price+0.84 (SEK/kWh) Export Revenue Elspot Price (SEK/kWh)

System Input Values

• Load profile (from Wallenstam AB )
• Hourly load profile of a residential building that comprises of 140 apartments in 2014;
• Located in Gothenburg;
• Weather profile (from Meteonorm)
• Global horizontal radiation (W/m2)
• Diffuse horizontal radiation (W/m2)
• Wind speed (m/s)
• Ambient temperature (°C)
• Provided by Meteonorm, generated from recorded weather data through a stochastic model
• Elspot Price (from Nordpool)
• Hourly Price in 2014
• SE3

• Elspot Price

System Input Values

Photovoltaic Model

• Five parameter single diode model

1

s

V IR s a l l sh

V IR I I I e R



          

V-I curve, P-I curve and MPPT

PV model

• Five parameter single diode model

Source: NREL System Advisory Database

PV model

V-I curve with different radiation and ambient temperature

Ambient Temperature, 25 °C Radiation, 100 W/m2

PV model

PV Production at Azimuth angle of 0° and Tilt Angle of 36°

• Azimuth and Tilt Angle

Battery Model

• Battery Voltage-Current Model
• Battery Lifetime Model
• Charge and discharge characteristics under different SOC and current
• Take into account the influence of SOC and current on the round trip efficiency
• V=f(SOC,I)
• Estimate the batter life time with the consideration of operation condition
• Lifetime=g(SOC)

Battery Voltage-Current Model

*

exp( ) Q Q V E K i K it A t iR Q it Q it         

  

*

exp( ) 0.1 Q Q V E K i K it A t iR Q it Q it         

  

Charge Discharge Lead Acid Battery Lithium ion Battery Charge and discharge performance of two batteries

Battery Lifetime Model

• Battery life cycle numbers and cycle depth

Lead Acid Battery LiPFO4 Battery DOD Cycle life DOD Cycle life 100% 250 80% 700 80% 300 50% 1440 50% 500 20% 2900 30% 1200 10% 6300

Battery C d m R-square @ 80% DOD Calender life Lithium ion 594.3

• 7.703e-12

1.022 0.9988 746 8Years Lead Acid 209.1

• 3.335e-10

1.448 0.9951 289 3Years 𝑂 = 𝐷 𝐸𝑃𝐸 − 𝑒 𝑛 Three Parameter N-S Curve Function

Battery Lifetime Model

It decompose the complex cycles to microcycles of different DOD

• Rainflow Counting Method

Cycle No. Cycle Path Cycle Depth Half or Full Cycle 1 ACDE 60% 0.5 2 EFGI 65% 0.5 3 BCD 20% 1 4 FGH 20% 1 Rainflow Decomposed Microcycles with Different DOD

𝑀𝑑𝑧𝑑𝑚𝑓 = 𝑂𝑡𝑢 𝑂𝑠𝑓𝑒 = 𝑂𝑡𝑢 𝐸𝑃𝐸𝑗 − 𝑒 𝐸𝑃𝐸𝑇𝑈 − 𝑒

𝑛

× 𝑆𝑗

Cycle lifetime

𝑀 = min 𝑀𝑑𝑧𝑑𝑚𝑓, 𝑀𝑑𝑏𝑚𝑓𝑜𝑒𝑏𝑠

Hydrogen Storage Model

• Power-Current Curve

SPE Electrolyzer PEM Fuel Cell H2 Storage

Power P-I Curve Current Hydrogen Power P-I Curve Current Hydrogen P-I curve for Fuel Cell and Electrolyzer

• Experimental Data
• Interpolate in the simulation

Genetic Algorithm

Generations 200 Population size 50 Algorithm Variant of NSGA II Crossover function Heuristic Crossover rate (%) 50 Mutation function Uniform Mutation rate (%) 5

Createa population

• f chromosomes

Determine the fitness of each individual Select next generation Termination Criterion? Perform reproduction using crossover Mutation Result PV Size Battery Size Electrolyzer Size Fitness Function NPV and SSR

Results and Discussion

• Comparison between battery storage and hydrogen storage
• Sensitivity Study of the component cost of the hydrogen storage system
• Hybrid battery and hydrogen storage system

Battery storage vs. Hydrogen storage

SSR vs. NPV of battery storage and hydrogen storage

Higher storage capacity

Hydrogen Storage Pareto Front from GA No Storage Battery Storage

Sensitivity study with hydrogen storage system

Electrolyzer Cost Fuel Cell Cost Hydrogen Tank Cost

PV Capacity 200 kWp

Hybrid battery and hydrogen storage system

Comparison between hybrid storage system with single battery or hydrogen storage system

Conclusion

• Battery storage system is superior to the hydrogen storage system since it

has higher NPV with the same achieved SSR;

• Electrolyzer cost is the most sensitive factor in the hydrogen storage

system for achieving higher SSR and NPV;

• Hybrid battery and hydrogen storage system can take advantages of both

individual system and achieve much better system performance.

Thank you! Tack!