Expansion and Operation Strategies in a Renewable and Hydro-Based - - PowerPoint PPT Presentation

Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System

4th IEEE and Cigré International Workshop "Hydro Scheduling in Competitive Markets", Radisson Blu Royal Hotel in Bergen, Norway, Bergen, June 14th - 15th, 2012 Egill Benedikt Hreinsson University of Iceland May 30, 2012

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 1 / 22

Outline

1 Introduction

The objectives of the paper Icelandic system overview with generation and resources

2 Generic model formulation 3 Modeling requirements and time frame

LTM and STM interaction Zones for different time scales Time scale decomposition STM time scale representation

4 Conclusions

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 2 / 22

Introduction The objectives of the paper

The objective of the paper

Give an overview of generation, load and resources in the small island system of ICELAND. Discuss new markets and export with wind integration. Present a generic hydro based system operations problem. Discuss model objectives and time frame (scale) requirements for this problem Draw some conclusions regarding modeling approaches to meet future requirements

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 3 / 22

Introduction Icelandic system overview with generation and resources

The Iceland Electrical Power System

Figure 1: xxx

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 4 / 22

Introduction Icelandic system overview with generation and resources

Review of generation

Total installed capacity in Iceland about 2.5 GW. Principal plants: Hydro: Kárahnjúkar (690 MW) Búrfell (270 MW) Hrauneyjafoss (210 MW) Blanda (150 MW) Sigalda (150 MW Sultartangi (120 MW) Sog (90 MW in 3 plants) Vatnsfell (65 MW) Andakíll (8 MW) Elliðaár (3 MW) Geothermal: Hellisheiði (303 MW) Nesjavellir (120 MW) Reykjanes (100 MW) Svartsengi (75 MW) Krafla (60 MW) Bjarnarflag (3 MW) Krafla (60 MW) Straumsvík (35 MW) (Gas turbine)

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 5 / 22

Introduction Icelandic system overview with generation and resources

Location of Hydro and Geothermal Projects

{Fig:rammi}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 6 / 22

Introduction Icelandic system overview with generation and resources

Geothermal stations

Nesjavallavirkjun Reykjanesvirkjun Kröflustöð Svartsengi

{Fig:geo}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 7 / 22

Introduction Icelandic system overview with generation and resources

Hydroelectric stations

Fljótsdalsstöð Vatnsfellsstöð Mjólkárvirkjun Lagarfossstöð

{Fig:hydro}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 8 / 22

Introduction Icelandic system overview with generation and resources

Dettifoss

{Fig:dettifoss}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 9 / 22

Introduction Icelandic system overview with generation and resources

Primary energy use in Iceland from 1940 to 2010

50 100 200 250

Hydro

100 60 80 20 40

1970 1990 2010

Fractional breakdown:

Primary energy utilization (PJ)

Hydro

Mór

Oil Oil Coal Peat Coal

150

Geothermal Geothermal

2000 1980 1960 1950

1950 1970 1990 2010

1940 Figure 2: Primary energy use in Iceland from 1940 to 2010 in PJ (PetaJoule) {Fig:r1}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 10 / 22

Introduction Icelandic system overview with generation and resources

Electrical energy sales of Landsvirkjun

14 12 Landsvirkjun Electricity sales (TWh/year) Energy Intensive Industry General Demand Landsnet 10 8 6 4 2 2011 2010 2005 2000 1995 1990 1985 1980 1975 1970 Elkem (1979) Alusuise (1969) Century Aluminium (1998) Alcoa (2007)

Figure 3: Electrical energy sales of Landsvirkjun; 1966 - 2011 {Fig:lva}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 11 / 22

Introduction Icelandic system overview with generation and resources

Submarine cable interconnections

k

1 2 5 k m 7 6 k m Norned, 580 km Iceland HVDC link options (Under consideration) North Sea Supergrid (Proposed) Existing LEGEND: 1 9 k m 1170 km

Figure 4: Iceland HVDC Cable Routes, NORNED and the North Sea Supergrid {Fig:hvdc}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 12 / 22

Generic model formulation

Model formulation

max

p,x,v,s f (p)

(1)

{eq:A1}

f is the objective function representing benefit, such as income minus cost, and p is a vector of generation and load variables. (1) is subject to the following constraints, where (2) are the water balance equations with hydraulic network topology and deterministic inflow series: gw(v, x, s) = 0 (2)

{eq:A2}

v is a vector of reservoir volumes, x is the release and s is spill in all periods in all reservoirs. The vector of Lagrange multipliers, Λw, with (2) are water values at each instant in each reservoir or (3), Λw =

• λw1,

λw2, · · · · · · T (3)

{eq:A2x}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 13 / 22

Generic model formulation

Load, network and market constraints

Equation (4) are the load, network and market constraints: (See (8) below) gL(p) = 0 (4)

{eq:A3}

Similarly, the vector of Lagrange multipliers, ΛL, associated with (4) are the shadow power prices at each instant at each node, or (5), ΛL =

• λL1,

λL2, · · · · · · T (5)

{eq:A3x}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 14 / 22

Generic model formulation

Technical generation constraints

The technical generation constraints are in (6), representing, for instance, the nonlinearities and head dependence in hydro stations: gt(p, x, v) = 0 (6)

{eq:A4}

Finally upper and lower bounds on all the variables are defined by (7): pmin p pmax vmin v vmax xmin x xmax s 0 (7)

{eq:A5}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 15 / 22

Generic model formulation

Subvector for generation/power flow

This above definition can be represented by sub-vectors: p =   pg ps pm   (8)

{eq:A6}

For all periods: pg is vector of generated power in hydro and thermal, (Wind is assumed a deterministic input) ps is a vector of the power flow in the electrical network, for instance using a DC/linear load flow representation (No voltage or phase angle). pm, is a vector of sold energy for instance on the spot market.

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 16 / 22

Modeling requirements and time frame LTM and STM interaction

Long and short term models

Short term model STM Long term model LTM

uS uL

Figure 5: LTM with time step of a ”week”and a horizon of years. STM with a time step of 30 minutes and a horzon of weeks. Interacting variables (vectors) are uL and uS {Fig:shortlong}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 17 / 22

Modeling requirements and time frame Zones for different time scales

A generic power system

Hydro (1) Hydro (2) Hydro (3) Other (5) "loads" Wind (4) Electrical transmission line LEGEND: Electrical substation/bus Water flow Electrical load/customers/market Time series inflow Electrical generator Reservoir with inflow Zone 2 with both STM and LTM given the presence of short term phenomena Zone 1 where short term phenomena such as daily reservoir fluctuations may be

• negligible. Therefore

LTM may suffice Electricity spot market or specific contracts/customers

Zone 2 Zone 1

xt,1 pg,t,1 pg,t,5 xt,2 pg,t,2 vt,1 xt,3 pg,t,3 vt,3 vt,2

{Fig:sys2}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 18 / 22

Modeling requirements and time frame Time scale decomposition

Time decomposition into LTM, MTM and STM

Mar Mar May Jul Jan 1st operational year Load/demand 1st operational month LTM STM 24 hours/1 day Nov Sep 1 3 30 28 26 24 Jan Nov Sep

Figure 6: Time decomposition into LTM, MTM and STM. A level below with very short variations (minutes) could be valuable for wind energy analysis {Fig:load1}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 19 / 22

Modeling requirements and time frame STM time scale representation

STM interaction and representation

1 C1 - Continuous Load duration curve (LDC) by sorting 8760

hours/year into an LDC, as 1 year’s distribution of load. Chronology is thereby removed, and is therefore not well suited to hydro systems with time interdependence.

2 L1 - Stepwise Long Term Load. Derive 12, 26 or 52 load values for a

time step in LTM of a week, up to a month.

3 LC2 An LDC added in the week to account for variations within the

week.

4 LS1. A complete chronology maintained in all periods of both the

LTM (weeks) and STM (hours).

5 LSC. Here we assume the merging of LC2 and LS1: (a) flat load, (b)

LDC with wind deducted and (c) total chronology

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 20 / 22

Conclusions

Conclusions and Discussion

Wind resources and spot markets seem likely to pose new interesting challenges to traditional hydro based system modeling and operations. LTM time scales have been used with good results for a hydro system without the wind resource. With the integration of wind and spot markets STM seems an important addition to interact with the new resources/markets and the LTM. The many possible ways of decomposing the time scale and interacting with data between the LTM and STM will affect the accuracy and computational efficiency and should be evaluated carefully with other modeling aspects and methodology. To address transmission constraints and losses in waterways, an STM seems an important aspect of the LTM modeling framework.

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 21 / 22

Conclusions

Thank you

{Fig:thank}

Egill Benedikt Hreinsson University of Iceland () Expansion and Operation Strategies in a Renewable and Hydro-Based Island Power System May 30, 2012 22 / 22