Advanced Gas Turbine and sCO2 Combined Cycle

Power System
UTSR Meeting
November 6, 2019
DOE Award Number: DE-FE0031619

Kevin Hoopes (SwRI)

Tim Allison (SwRI)
Michael Marshall (SwRI)

Karl Wygant (Hanwha Power Systems)

Jonathan Bygrave (Hanwha Power Systems)
Robert Pelton (Hanwha Power Systems)

Chris Lyons (Solar Turbines)

Scott Schubring (Williams)

Project background and motivation: Large scale combined cycles

Modern gas turbines are highly efficient with

thermal efficiencies between 30 and 45%.

To drive overall efficiency even higher, large gas

turbine power plants use a Waste Heat Recovery
System (WHRS), also called a bottoming cycle, to
extract otherwise wasted heat from the gas turbine
exhaust.

This combination of primary and bottoming cycles

is called a Combined Cycle Gas Turbine or CCGT.

For large CCGT plants a steam Rankine WHRS is

typically used. The addition of this WHRS allows for
Schematic of large scale Combined Cycle Gas Turbine power plant from Siemens. overall plant thermal efficiency to reach nearly 63%
Large scale plants of this type can be upwards of 300 MWe.
in modern, large, utility scale plants.

2
Project background and motivation: sCO2 small scale combined cycle
WHR systems are not typically added to smaller gas
turbines due to:
• Size
• Complexity
• Initial capital cost
• On site water requirements

A WHR system based on super critical CO2 (sCO2) as a

working fluid could address all of these issues.

This project will develop an advanced WHR cycle using

sCO2 as a working fluid and create a conceptual design
of a complete WHR system applicable to existing gas
turbine installations. This WHRS will be:
Solar Turbines gas turbines and gas compressor at a Spectra Energy • Highly efficient
natural gas compression station.
• Modular and skidable
• Compatible with air cooling
• Allow for advanced load following

3
WHRS Development Technical Path

WHR development used the following technical path:

1. Choose target Gas Turbine

2. Choose cycle configuration
3. Optimize cycle component operating points
4. Conceptual cycle component design

4
Project background: Integrally-Geared turbomachinery

In order to further increase efficiency and compactness,

the WHRS will use integrally-geared turbomachinery.

An integrally-geared machine consists of a central bull

gear connected to one or more pinion shafts which
contain one or two impellers each. These impellers can
be radial compressors or radial turbines.

Integrally-geared machines have the following benefits:

• Independent pinion speeds increase overall machine
efficiency by allowing optimized aero matching
• Easy access to the process gas between stages for
intercooling and or reheating
• Stage access allows for each stage to have variable
IGVs for better control
• Compactness
Single pinion integrally-geared compressor from Hanwha Power Systems

5
Solar Turbines Titan 130
The Solar Turbines Titan 130 was chosen as the target
gas turbine because:

1) The size of the Titan 130 is ideal for applying a

sCO2 WHRS that uses integrally geared
turbomachinery
2) There are over 850 Titan 130s installed around
the world.
3) Solar Turbines has confirmed that there is
significant demand for this size of gas turbine and
this is likely to continue.
4) While larger gas turbines are generally more
thermally efficient, there is a smaller installed
base and so the numbers of WHRSs required will
be less.
5) The Titan 130 has been tested with alternative
Schematic of a Solar Turbines Titan 130 which produces 16.5 MWe fuels such as coal derived syngas.

6
Preheat Cycle
Preliminary analysis showed that the preheat cycle is an
excellent choice for Waste Heat Recovery applications as it
allows for good recuperation along with high energy extraction
from the waste heat stream while minimizing complexity.

If the cycle is modified by splitting the recuperator it is suitable

for a wide range of gas turbine fuels (natural gas, coal derived
syngas, hydrogen, etc.) which have different acid dew point
temperatures. This could take the form of two separate
Preheat cycle
recuperators or a monolithic recuperator with an additional side
stream outlet.

Preheat cycle with split recuperator (PreheatSR)

Typical PreheatSR cycle Ph/Ts diagram

7
Acid Dew Point Corrosion
Gas turbine exhaust can contain Sulphur and other contaminants, if
the temperature of the stream drops too low these can condense into
liquid acid droplets and cause corrosion in the coldest end of the flue
gas / CO2 heat exchanger (cold CO2 inlet). HRSGs are also susceptible
to this.

For “clean” pipeline natural gas, 60[C] is a standard industry limit for
Typical HRSG acid corrosion (from NETRA)
HRSGs. For coal derived syngas the acid dew point temperature rises
to above 100[C] which limits the extractable heat from the waste heat
stream.

The preheat cycle with a split recuperator can be configured for any
Heater Inlet Temperature (HIT) limit and is therefore adaptable to any
gas turbine fuel. There is a loss in performance, but it is small (2.5%
for 100[C] limit).

With the split recuperator the cycle outperforms the standard

preheat cycle without an HIT and does much better with an HIT
imposed Comparison of standard Preheat and Preheat with split
recuperator (PreheatSR)
8
 Cycle Optimization

With a cycle selected, the cycle component operating

points / design points must be selected.

• What ambient temperature should we use for the

cycle design point?
• Where should we place the design point for optimal
operation throughout the year?
• What approach temperatures/effectiveness should
we choose for the cooler, heater, recuperators?

To address this, we used a multi-objective optimization PreheatSR Cycle

routine to explore the cycle design space.

The goal of the optimization was to produce a series of

designs to maximize cycle performance while minimizing
cycle capital cost.

9
Optimization Routine
• The goal is to produce a Pareto front of capital cost vs cycle performance to select a cycle design point.
• Two nested optimization routines are used. The outer optimization works with design parameters while
the inner routine optimizes how the machine is run throughout a year at off-design ambient conditions.

Outer Optimization
Inner Fitness Function
Givens: Cycle layout, operating site temp profile
Optimizes cycle performance for a series of ambient temperatures then
Objectives: Minimize cost, maximize energy out
integrates that performance to get overall yearly energy out. Also calculates
Constraints: CO2 > 60 C into Heater HX inlet (acid)
CAPEX from component cost correlations.
Fitness function input variables:
Givens: Cycle design variables, GT operating profile, site ambient T profile
• Design low pressure
Objectives: Maximize cycle performance for series of ambient temperatures
• Design high pressure
Constraints: CO2 above critical P and T at compressor inlet.
• Design heater eff.
Inner fitness function input variables for each ambient T:
• Design ambient air T
• Cooler air flow rate
• Design cooler eff.
• Flow split
• Design recuperator eff.
• Cycle massflow
• Design cycle massflow
Inner fitness function output variables at each ambient T:
• Design flow split
• Energy out
• Design air massflow
Inner Fitness Function Output
Fitness function output variables:
• CAPEX cost from cost correlations
• Configuration cost
• Integrated yearly performance from series of ambient T optimizations
• Yearly energy out

10
 Outer fitness function details

The inner fitness function follows the following procedure:

1. Obtain design variables to evaluate from outer optimizer

2. Choose series of ambient temperatures (6)
3. Optimize operating points at each ambient temperature to maximize power
output using off-design models for cycle components
• Vary cycle massflow, cooler air flow, and flow split
• Find best combination which yields max power at each ambient temperature
4. Create net power out vs ambient temperature curve from optimized points
5. Interpolate net power out vs temperature and site ambient temperature data to
create power vs time for the whole year at the target site
6. Integrate power vs time to get net energy produced for the entire year
7. Compute capex cost using cost functions for the set of design variables
8. Return net energy and capex to outer optimizer For this method, off-design
performance models and cost
models are needed for each
component.
 Component Cost/Off-Design Models

To complete the optimization, cost and off-design models are needed

for each component in the cycle.

A recent paper from NETL has cost models for heaters, coolers,
recuperators, and turbomachinery. The heater model is not directly
applicable as it is for a natural gas fired direct heater. We worked
with HRSG vendors to get a more appropriate heater cost model. We
also used a custom model from Hanwha for turbomachinery costs.

Component Cost Model Off-Design Model

Turbomachinery Hanwha custom model Hanwha maps
Heater SwRI Correlation from Quotes Conductance Ratio Method: CR from SunShot
Heater/Quotes
Recuperator NETL Conductance Ratio Method: CR from SunShot
Recuperator
Cooler NETL Conductance Ratio Method: CR from Quotes
Gas Turbine N/A Data from Solar

Recuperator Model Comparison

 Conductance Ratio Method

Recuperator, heater, and cooler On-Design

use the conductance ratio 1 1 1 ℎ𝐴𝐴ℎ𝑜𝑜𝑜𝑜
method for off-design 𝑄𝑄 = 𝑈𝑈𝑈𝑈 · 𝑓𝑓 ∆𝑇𝑇 = � + 𝑅𝑅𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 + � ℎ𝐴𝐴 ∗
=
𝑈𝑈𝑈𝑈 ℎ𝐴𝐴 ℎ𝑜𝑜𝑜𝑜 ℎ𝐴𝐴 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 ℎ𝐴𝐴𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
performance.
Calculate overall Use ℎ𝐴𝐴 ∗ to divide UA
This method has been validated resistance required into constituent terms
against multiple experimental
data sets (3 published) including Scale with off-design conditions
several sCO2 heat exchangers
(GT2017-64908). Off-Design 1 1 1
Combine scaled hA terms to form new UA = � + 𝑅𝑅𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 + �
𝑈𝑈𝑈𝑈 ℎ𝐴𝐴 ℎ𝑜𝑜𝑜𝑜 ℎ𝐴𝐴 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
Essentially the method scales on
design performance using the 𝑄𝑄 = 𝑈𝑈𝑈𝑈 · 𝑓𝑓 ∆𝑇𝑇 = 𝑚𝑚̇ Δℎℎ𝑜𝑜𝑜𝑜 = 𝑚𝑚̇ Δℎ𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
off-design flow conditions.
Use UA term and energy conservation to predict outlet temperatures

13
Calculation of Yearly Energy Output
• Calculation of a yearly energy output for a specific design uses two components: the hourly temperature
variation for a calendar year, and the net output power vs temperature relationship.
• The net power output at each ambient data point uses cubic interpolation from the power vs temperature
curve, and Simpson’s rule is used to integrate the net power output over a year’s time.

Yearly ambient temperature plot for weather station in WV Power vs Ambient temperature curve
(Site of prospective WHRU at gas processing plant)

14
Optimization Results

Multi-objective optimization result Breakout of component costs

15
Optimization Results

After the application of a discount rate,

net present value allows for the
selection of the overall most profitable
cycle.

Adjustments need to be added for:

• electricity price variation
• Gas turbine load variation

Net Present Value for a series of discount rates

16
Cycle Details

PreheatSR Cycle

Design Parameters
Compressor Inlet Pressure 89.3 [bar]
Compressor Discharge Pressure 252.7 [bar]
Turbine Inlet Temperature 344.8 [C]
Ambient Temperature 17 [C]
Cycle D design point temperature vs yearly
ambient temperature variation
17
Titan 130 Combined Cycle Efficiency

At ISO conditions in an electrical generation

configuration, the Titan 130 generates 16.530 [MWe]
at a thermal efficiency of 35.4%

At ISO conditions and 100% load, the WHRU cycle

adds an additional 5.743 [MWe] which creates an
overall combined cycle efficiency of 47.7%

Traditional HRSGs can add up to 3.4 [MWe]. The

developed sCO2 WHRU provides a 68% improvement
over traditional HRSGs.

Schematic of a Solar Turbines Titan 130

18
Powerblock Skid
Recuprator
Conceptual design included considering how the
components would fit on a skid as well as conceptual
Core Assembly

turbomachinery sizing including preliminary aerodynamics

and rotordyanmics. Detailed engineering is needed to flesh Generator

out the design before manufacturing.

Oil Cooler

VFD

Oil Reservoir

Control Panel
Startup
Seal Rack
Motor

Recovery
Compressor

Conceptual skid layout

Typical rotor layout

19
sCO2 Power System – Core Configuration
Turbine & Compressor Pinions High and Low Pressure Pinions

Generator Speed: 1800 RPM Generator Speed: 3600 RPM Generator Speed: 1800 RPM Generator Speed: 3600 RPM
Gear CD: 1x Gear CD: 0.6x Gear CD: 0.9x Gear CD: 0.5x
Mechanical Losses: 100% Mechanical Losses: 95% Mechanical Losses: 58% Mechanical Losses: 56%

• Two generator options were considered

1. 1800 rpm generator: Lowest cost
2. 3600rpm generator: Smallest package
• Two different turbomachinery configurations were evaluated.
1. Turbine/Comp Pinions: Turbines and compressors on separate pinions
2. HP/LP Pinions: HP stages (Turbine S1 and Comp. S2) and LP stages (Turbine S2 and comp. S1) on separate pinions
• An HP/LP configuration reduces the size of the gearbox and the mechanical losses
• A 3600rpm generator is not a common offering and is much more expensive at powers below 10MWe
Preliminary Aero/Rotordyanmic Design

Conceptual and preliminary design work is

underway on the turbomachinery block
components including aerodynamic and
rotordynamic models.

Expander preliminary layout / geometry

Preliminary Expander CFD

Geometry used in the rotordynamic model of LP Pinion

21
Summary
By utilizing a multi-objective optimization framework, this project
has developed a conceptual design for a sCO2 WHR System for a
Solar Turbines Titan 130. The chosen conceptual sCO2 cycle
increases the thermodynamic efficiency from 35.4% to 47.7%

Recuprator

Core Assembly

Generator

Oil Cooler

VFD
Oil Reservoir

Control Panel
Startup
Seal Rack
Motor
Recovery
Compressor

22