LVDC grid based on PV energy sources and multiple electrochemical - - PowerPoint PPT Presentation

LVDC grid based on PV energy sources and multiple electrochemical storage technologies

Kolja NEUHAUS, Jeremy DULOUT, Corinne ALONSO

Université Paul Sabatier LAAS-CNRS Toulouse

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

Summary

▪ Energy grids and their evolution to smart grids ▪ LVDC grid, NeoCampus context ▪ The ADREAM test platform - energy optimized building ▪ Cluster analysis of production and consumption ▪ Storage units adapted for energy optimized buildings ▪ Conclusion

2

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

Energy grids and their evolution to smart grids

3

▪ Vertical grid structure

◆ Main focus on security of supply (oil crisis 1973) ◆ Matching load to the production ◆ Unidirectional transport from producer to consumer

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

Energy grids and their evolution to smart grids

4

▪ Horizontal grid structure and smart grid

◆ Every element except grid operation is subject to competition since the years 2000’s. ◆ Free choice of energy provider and self-production/consumption made possible. ◆ Ability to integrate intermittent renewable and distributed production. ◆ Multidirectional transport based on energy flow mitigation with storage.

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

The ADREAM test platform - energy optimized building

5

• Total photovoltaic surface:

720 m²

• Total peak power: 100 kWp
• 4 different areas with

different inclinations

• Typical PV production data

for an energy optimized building.

Typical consumption data for a three phased AC micro-grid for energy optimized building.

▪ George Giralt building (ADREAM)

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

▪ Solar radiation and PV production

6

The ADREAM test Platform - energy optimized building

◆ Good knowledge of the producible means good knowledge of the production. ◆ Direct impact of solar intermittency on PV production and storage strategies. ◆ Solar radiation study is used as high precision comparison and verification of PV performance.

Solar irradiation sensors on their inclinable stand Pyranometer doted with a solar ring to measure diffuse and reflected irradiation

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

Cluster analysis of production and consumption

7

▪ Lots of data on photovoltaic energy production

Daily production data with one sample per minute

▪ Lots of data of building energy consumption

Daily consumption data with one sample per minute

▪ High intermittency is a determinant factor

Solar radiation, building occupation and exterior conditions

▪ How to determinate simple and meaningful daily profiles from all this data ?

Used to study storage strategies adapted to the intermittent production and consumption

K-medoids Clustering

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

Cluster analysis of production and consumption

8

Raw data Choice of K (number of clusters) + first medoids Calculating distance between data point and medoids Allocating data point to cluster based on minimal distance with medoid Recalculating medoids for each cluster Medoidsi == Medoidsi-1

YES NO

K clusters + K medoids

2 3 4 1 5

The algorithm is initialized with the raw

• data. The number of desired clusters (K) is

chosen by the user and one data vector is chosen as medoid for each cluster. The distance between each data point and the medoids is calculated. Each data is assigned to the closest medoid, forming the clusters. The new medoid of each cluster is selected. The algorithm stops when the K medoids have not changed between two iterations.

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS 9

Cluster analysis of production and consumption

Time (hh:mm)

00:00 02:24 04:48 07:12 09:36 12:00 14:24 16:48 19:12 21:36 00:00

PV Power ( W)

10 4

• 6
• 5
• 4
• 3
• 2
• 1

1

Low intermittent radiation Low uniform radiation High unifirm radiation High intermittent radiation

PV Production clustering results 2015

Time (hh:mm)

00:00 02:24 04:48 07:12 09:36 12:00 14:24 16:48 19:12 21:36 00:00

Consumed Power

(W)

104 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 Day type 1 Day type 2 Day type 3

Building consumption clustering results 2015

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

Storage units adapted for energy optimized buildings

10

Important parameters for durable stationary energy projects :

• Lifetime (depending on Depth Of Discharge per cycle)
• Efficiency (ratio of charge/discharged energy per cycle)
• Total environmental footprint
• Price
• Specific energy > Specific Power

Selected technologies :

Lead-Acid batteries Lithium iron phosphate batteries Lithium supercapacitors and hybrids

▪ Relatively low specific energy ▪ Lowest price ▪ Jellified electrolyte (AGM, OPzV) for low maintenance ▪ 500-1500 cycle lifetime ▪ Low self-discharge ▪ Good specific energy ▪ Medium price ▪ Close to no maintenance ▪ 2500-3000 cycle lifetime ▪ Low environmental footprint ▪ Good specific energy AND power (good for intermittency) ▪ High price ▪ Close to no maintenance ▪ High cycle lifetime (>5000) ▪ Low environmental footprint

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

Conclusion

▪ To conceive a local optimized energy grid based on photovoltaic sources, a good knowledge of the solar producible, its intermittence and its production is needed. ▪ Production and consumption must be studied in order to develop adequate energy and storage managing for optimal energy flow and minimize losses. ▪ Lead-acid AGM, OPzV and lithium iron phosphate batteries are particularly adapted to stationary renewable energy production structures. Electrochemical hybrid LIC has interesting technical specifications and seems promising. ▪ One key point in energy transfer optimization is a better knowledge of storage units. For this purpose, a model of state of charge (SOC) and state of health (SOH) in battery storage units is necessary and is in development stage for the next step of this work.

11

Laboratoire d’analyse et d’architecture des systèmes du CNRS Laboratoire d’analyse et d’architecture des systèmes du CNRS

Thank you for your attention

12