Ngawha Reservoir Model - JOGMECgeothermal.jogmec.go.jp/report/file/session_160602_06.pdfNgawha...

GNS Science

Ngawha Reservoir Model

John Burnell

JOGMEC-GNS Workshop, Tokyo

2 June 2016

GNS Science

Outline

• Purpose

• Conceptual Model

• Model Structure

• Model Verification

• Expansion Scenario

• Surface Features Model

Slide 2

GNS Science

Purpose

• To develop a numerical reservoir model of the Ngawha Geothermal Field for consenting purposes

• Objectives

– Obtain an acceptable match to historical data

– Use a dual porosity model to more accurately model

injection returns in the reservoir

– Consider effects of possible future expansion

– Predict impact on surface features

Slide 3

GNS Science

Conceptual Model

Slide 4

GNS Science

Reservoir Characteristics

• Reservoir hosted in Waipapa Greywacke between 500m and 1500m depth

– Low porosity

– Permeability due to fractures and faults

• Low permeability siltstones form a cap

– Some transport through faults and fractures

Slide 5

GNS Science

Resource Areas

Operations area

Slide 6

GNS Science

Reservoir Characteristics II

• Geology

– Permeable greywacke

– Cap with small channels connecting to surface

• Temperatures in permeable reservoir are ~230 °C

• Hotter temperatures in the north

• Pre-production testing suggested fractured

reservoir with high permeability

• Hot upflow > 300°C, located in north < 10 kg/s

• CO2 concentrations ~1.5%

Slide 7

GNS Science

Hydrological Model

Slide 8

GNS Science

Model Validation Data

Slide 9

• Pre-production temperature and pressure measurements.

• Measured changes in reservoir pressure due to production.

• Changes in production enthalpies.

• Changes in CO2 concentrations in production wells.

• Measured pressure changes in the 1983

interference tests.

• Travel times from the 2011 tracer test.

GNS Science

Dual Porosity Numerical Reservoir Model

Slide 10

GNS Science

Dual Porosity Model

Matrix

Fracture

Used to represent flows in fractures• Fast flow paths• Small volumes

• Model cooling

Slide 11

GNS Science

Model Grid

Slide 12

GNS Science

Geology and Resource Boundary

Permeable inner region

Slide 13

GNS Science

Upflow

Upflow = 10 kg/s

Slide 14

GNS Science

Model Verification

Slide 15

GNS Science

Methodology

• Run natural state model

– Compare to measured temperatures and pressures

• Run historical production

– Compare to measured pressure changes, enthalpies and CO2 concentrations

• Add tracer injection.

– Compare to measured tracer

• Run production and injection for 1983

interference test

– Compare to measured pressures

Slide 16

GNS Science

Natural State Temperatures

Slide 17

GNS ScienceSlide 18

GNS Science

Some Improvements Still Needed

Slide 19

GNS Science

Natural State Pressures

Slide 20

GNS Science

Changes Due to Development

Slide 21

GNS ScienceSlide 22

GNS Science

Pressures in Injectors Don't Match as Well

Slide 23

GNS Science

Production Enthalpies

Slide 24

GNS ScienceSlide 25

GNS ScienceSlide 26

GNS Science

Issues with Enthalpies

• Inconsistent with measured tempertures

– NG9 temperatures give enthalpies of 980 kJ/kg

– Initial data – 920 kJ/kg

– Also inconsistency between well enthalpies and station enthalpies

– In other fields it is suggested that measurement uncertainties are ~100 kJ/kg

– If this is the case then the model agrees within

measurement uncertainty.

Slide 27

GNS ScienceSlide 28

GNS Science

CO2 Concentration

• CO2 gives information about long term connection between injectors and producers

Slide 29

GNS ScienceSlide 30

GNS ScienceSlide 31

GNS Science

Tracer Test – Injection into NG16

• Arrival times give information about short term transport between injectors and producers

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

Co

nc

en

tra

tio

n (

ug

/L)

NG16 Tracer

NG4

NG9

NG12

Slide 32

GNS Science

Model Results

Arrival time

Slide 33

GNS ScienceSlide 34

GNS Science

Interference Test – October 1983

• Pressure propagation across the reservoir

Slide 35

GNS Science

Overall Match

• Ngawha Numerical Reservoir Model is well-calibrated

• Matches data from many different aspects of the field

• Based on physical understanding

– Reduces uncertainty

Slide 36

GNS Science

Expansion Scenarios

Slide 37

GNS Science

Expansion Scenarios

• Two scenarios

– Scenario A: 25 MWe expansion

– Scenario B: 50 MWe expansion

• 25 MWe stage from 2020

• 50 MWe stage from 2023

• Adjusted injection rates to match observations

– Loss of gases (currently 0.5%)

– Other loses/spillages (0.35%)

• Used supplemental injection to control reservoir pressures

Slide 38

GNS Science

Scenario Pressures

Slide 39

GNS Science

Some pressure changes

Slide 40

GNS Science

Enthalpies Decline

Slide 41

GNS Science

Will need some plant changes

Slide 42

GNS Science

Surface Feature Modelling

Slide 43

GNS Science

Surface Feature Modelling

• Ngawha Springs are valued highly by local community

• Ngawha Springs have low flow rates

• Reservoir model is too coarse to capture changes in flow to surface features

• Developed a separate model of the area around the surface features

Slide 44

GNS Science

Schematic of Surface Feature Model

Deep Reservoir

Surface Waters

CaprockPermeable

Conduit

Slide 45

GNS Science

Model Grid

Slide 46

GNS Science

Vertical Extent

Slide 47

GNS Science

Natural State Temperatures

Slide 48

GNS Science

Natural State Pressures

Slide 49

GNS Science

Surface Feature Model Setup

• Fixed pressures and temperatures in the bottom boundary

– Corresponding to measured reservoir data

• Fixed conditions in atmosphere

• Flow driven by boundary conditions

– Adjust conduit permeability to match flow

– Modelled flow was 1.3 kg/s

Slide 50

GNS Science

Response to Production

• Used output from numerical reservoir model expansion scenario

– Reduced reservoir pressure by 0.3 bar

– Reduced reservoir temperature by 1 °C

– Reduced reservoir CO2 mass fraction from 1.3% to

0.1%

Slide 51

GNS Science

Surface Feature Model Results – Flow to

Surface

Slide 52

GNS Science

Surface Feature Model Results – Steam

Fraction

Slide 53

GNS Science

Surface Feature Model Results – Temperature

Slide 54

GNS Science

Surface Feature Model Results – CO2 in

Discharge

Slide 55

GNS Science

Chemistry Calculation for Surface Features

• Ran a calculation with GeoChemist Workbench to obtain spring chemistry

• Reduced the CO2 content and re-calculated

• The main changes are an increase in pH (6.2 => 6.95) and a reduction in HCO3-.

Slide 56

GNS Science

Conclusions

• Well-calibrated model for consenting purposes

• Matches many different aspects of flow in the

reservoir

• Pressures can be controlled for 50 MWeexpansion scenario

• Impact on surface features is minor

Slide 57