New Combined Cycle Technology For Achieving Ultra-Low Emissions

William S. Rollins, President, NovelEdge Technologies, LLC

Marty Fry, Applications Manager, Energy Systems Coen Company, Inc.
Marty Nygard, Principal Systems Engineer, Nooter/Eriksen

INTRODUCTION:

Today’s gas fired combined cycle plants are considered to be the most efficient and environmentally friendly means to produce electric power.   Yet environmental demands continually force developers to reduce emissions even further.   The increase in capital investment and operating costs required to meet future demands for lower emissions are expected to weigh heavily on the economics and funding required to develop new power plants.

However, there may now exist opportunities to significantly reduce the combustible emissions generated by gas turbines (GT) in combined cycle applications without appreciably adding to the cost of the project.

This paper offers the accumulative results from two separate technology studies: NovelEdgeTM Technology and thermal oxidation.  Utilizing the combination of technologies, this paper offers a theoretical optimization of a combined cycle system to achieve ultra low emissions.  The goal of the optimization was to develop a theoretical system to reduce the total combustible emissions generated by a typical three-on-one unfired combined cycle plant by 67%.  This technology is a result of the combined efforts of NovelEdge Technologies, Coen Company, and Nooter/Eriksen.

Through design, theoretical analysis, and computer modeling, an alternative method for reducing combustible emissions has been derived which promotes the thermal oxidation of residual combustible emissions from the gas turbine (GT) engine by utilizing supplemental gas fired duct burners.  This method for combustible emissions reduction is patent pending.  Although burners have traditionally contributed to emissions levels when employed in a combined cycle power plant, this optimized duct burning approach, when properly configured and operated, provides for a reduction of the combustible emissions from the GT and results in an ultra-low emission combined cycle power plant.

EMISSION CLASSIFICATIONS:

There are numerous emissions associated with combined cycle power plants.  Most of these emissions are a result of the combustion process, either in the gas turbine engine itself, or in the duct burners that are typically located in the HRSG.  When burning a hydrocarbon-based fuel, the products of combustion include carbon dioxide (CO2), water vapor (H2O), and other products, some of which are designated as criteria pollutants.

Some of the criteria pollutants that are monitored by the regulatory agencies, including the U.S. EPA and the individual state environmental agencies, include:


  • NOx – the oxides of nitrogen, i.e. NO, NO2, etc.
  • CO – carbon monoxide
  • UBHC – unburned hydrocarbons
  • VOC’s – volatile organic compounds
  • Particulates – small particles of unburned carbon, inert material, and sulfur based compounds.

Throughout this paper reference is made to combustible emissions. Combustible pollutants are defined as carbon monoxide (CO), unburned hydrocarbons (UBHC), volatile organic compounds (VOC’s), and carbon particulate. For the purposes of this paper we have assumed a weight percentage of 15% for the carbon content of the particulate emission from the GT.

Depending upon the GT combustor design, GT load point, fuel quality, and other factors, the rate of production in the GT combustion section for these criteria pollutants varies widely.  Typically, each GT will come from the manufacturer with a guarantee on the level of emissions that its engine produces.  For evaluation purposes, the guaranteed emission values representative of a typical GE Frame 7FA GT have been used as the base line for the turbine emissions.


BACKGROUND:

NovelEdgeTM Technology1 is a patented combined cycle process that utilizes a single pressure level HRSG in combination with continuous supplemental firing and the use of a large economizer section in the HRSG.  This results in a highly efficient, yet low cost combined cycle process.  In essence, this unique cycle replaces the heat recovery function of intermediate pressure (IP) and low pressure (LP) steam that is utilized in a conventional combined cycle with greater water flows and a large economizer in the exhaust end of the HRSG.  The use of supplemental firing is essential to efficient operation in the cycle, and the HRSG stack temperatures are comparable to those of the conventional cycle.

Figure 1 illustrates a typical process schematic for a conventional combined cycle.  Note the use of an HP, IP, and LP evaporator section in the HRSG.  A typical stack temperature for this conventional design without supplemental firing is 190° F.  Without the IP and LP steam production, the HRSG stack temperature in the unfired condition would be nominally 380° F.

Figure 2 illustrates the process schematic for the patented combined cycle.  Note here there is only one evaporator in the HRSG, the HP section, and there are two economizer sections shown.  An integral part of this design is the continuous use of supplemental firing in the operational range.  This modulated rate of firing below 100% GT load creates additional steam, which becomes the necessary feedwater flow in the economizer to reduce the GT exhaust gas temperature to the same 190° F at the stack of the HRSG as in the conventional plant.

Therefore, this new technology provides for the generous use of supplemental firing, which is well known in the industry to add low cost capacity.  However, due to improvements in the bottoming (steam) cycle, the typical degradation in heat rate that is incurred from supplemental firing is mitigated by the higher efficiency in the steam cycle.  This higher steam cycle efficiency comes from 3 major factors:


  1. All steam is at high pressure, no IP and LP steam is produced
  2. The HP steam pressure is typically higher than the conventional HP steam pressure, yielding a more efficient cycle
  3. Extraction steam feedwater heating is employed

The result from utilizing this newly patented technology can be seen in Figure 3.  This curve illustrates a performance comparison between various conventional plants and newly patented combined cycle plants utilizing “F” class GT engines in 1×1, 2×1, and 3×1 configurations.  For this new technology, these curves are for the moderate fired or non-water-wall HRSG applications (note that with water-wall HRSG construction, this new technology can produce up to 950 MW in a 2×1 configuration).  As can be seen from Figure 3, the new technology 2×1 produces up to 800 MW, versus only 600 for the conventional plant.  This represents a 33% increase in capacity.  The best heat rate for the new technology 2×1 facility is at a power output of slightly more than 600 MW, and the heat rate is within approximately 1% of the conventional plant’s best heat rate.

Since the patented combined cycle 2×1 plant can produce up to 800 MW, it can replace the conventional 3×1 plant with the same output, or similarly, the new technology 3×1 can replace the conventional 1200 MW plant, which is essentially 2 blocks of a 2×1 arrangement.  This means one less gas turbine, one less HRSG, and one less steam turbine is required to produce 1200 MW, which equates to a large savings in up front capital.

Since unfired conventional plants are considered to be the “cleanest” combined cycle plants, or in other words, the plants with the lowest specific emissions in tons/MWH (tons per megawatt-hour), then we will compare emissions for a new patented combined cycle technology 2×1 facility at 750 MW, versus a 3×1 conventional plant that has no supplemental firing and also produces 750 MW.  Note that this is comparing two plants of equal output, and it is comparing the continuously fired technology facility to the “best available specific emission” plant currently in the industry.

Fig 1, Conventional Combined Cycle

Fig 2, NovelEdgeTM Heat Balance

Fig 3 NovelEdgeTM to Conventional Performance Comparison – Moderate Firing


NOOTER/ERIKSEN HRSG:

Since this new combined cycle requires a distinctly new HRSG, NovelEdge Technologies has partnered with Nooter/Eriksen for the design and manufacture of the HRSG.  As discussed, this HRSG has only one drum, an HP drum, and a large economizer section.  In addition, it is designed with a duct burner system that operates on a continuous basis in the load range (typically 50% GT output to full plant load).

For a comparison to conventional emissions, an unfired 3×1 with “F” class GTs at 750 MW is being compared to a new technology application with a 2×1 arrangement at 750 MW.  The new technology HRSG for this application is shown in Figure 4.  Again note there is only one drum and a large economizer with a continuously fired duct burner.  The overall size for this HRSG is similar to the size of the conventional unfired HRSG.

NovelEdgeTM CYCLE ADVANTAGE:

The NovelEdgeTM cycle, the newly patented combined cycle, offers distinct advantages to power plant owners through its lower cost of construction and lower cost of operation.  The cycle is also uniquely suited to the emission incineration aspects as well.

This economic and low admissions advantage is a result of the high power density design for the cycle.  This is relative to the amount of power that is produced in the combined cycle per GT utilized.  This factor is important, because more GT engines in a cycle equate to more exhaust gas that can potentially contain emissions.

As seen in Figure 3, the new technology 2×1 arrangement, with only moderate firing, can produce up to 800 MW.  With a conventional combined cycle with similar “F” class gas turbines, three GT’s would be required.  Therefore, if the emissions levels of the duct burners are low, there is still a net reduction in emissions as there are now fewer GT engines that are producing emissions.

Let’s consider an example where the duct burners produce no emissions, and compare it to a state-of-the-art low emissions conventional cycle with no duct firing.  In this example, a 750 MW 3×1 conventional combined cycle would emit 1.50 times the emissions of this new technology 750 MW plant in a 2×1 arrangement.  This is quite simply a function of the number of GTs, given that the duct burners produce zero emissions.

When the effects of thermal oxidation are added, the emissions reductions become even more pronounced.  In this scenario, the duct burners reduce the combustible emissions that are emanating from the GT engines.  This now equates to not only lower emissions due to fewer HRSG stacks, but each stack will have lower emissions than its conventional combined cycle counterpart.  Therefore, there are fewer stacks, each with lower emissions, which result in a tremendous reduction in combustible emissions for a combined cycle power plant of equivalent power output.  With a 50% incineration rate, the plant combustible emissions for this new combined cycle are reduced by a factor of 3.

Therefore, the NovelEdgeTM cycle can contribute to the reduction of emissions merely by its virtue of being a high-power density design.  When the effects of thermal oxidation  are considered, it will only serve to further reduce the emissions already lowered by this phenomenon.


Figure 4, HRSG For F Type GT


PRINCIPLES OF COMBUSTION:

For combustion, the essential ingredients are sometimes referred to as the 3 – T’s of combustion; Time, Temperature, and Turbulence.  If these three factors are all correct, it is possible to have a combustion process that creates very few combustible emissions, assuming there exists adequate oxygen for the reactions to take place.

Temperature, the first factor to be discussed, is essentially just as its name implies.  If the fuel/air mixture (in our case, fuel and turbine exhaust gas, TEG) does not reach an adequate temperature, it will not ignite.  Therefore, it is imperative that the combustion reactants and oxygen source attain sufficient temperature for combustion to initiate (and this includes the hydrocarbon emissions from the GT engine).  Thus, the firing temperatures in the duct burners must be above the ignition temperatures for the combustion products.  Table 1 provides a list of ignition temperatures (in air) for some of the reactants found in the combustion process.


Table 1

Reactant

Ignition Temperature

Carbon Monoxide (CO)

1128

Methane (CH4)

1170

Ethane (C2H6)

882

Propane (C­3H8)

871


Turbulence is a general term, and in this application it refers to several factors for adequate combustion in a duct burner.  First, in order for combustion to take place, there must be an adequate amount of oxygen in the fuel/air mixture to support the process.  In addition, this necessary oxygen must be evenly dispersed.  Up stream of the duct burner, the mechanism that provides an even distribution of turbine exhaust gas (i.e. Oxygen) to the duct burner is called distribution grid.  It is normally the responsibility of the HRSG manufacturer to ensure that adequate distribution is provided to the burner.

Secondly, Turbulence is also function of the duct burner itself.  Although there may be adequate distribution to the duct burner, the exhaust gas and fuel need to be mixed by the duct burner.  It is essential that this mixing process be very efficient, otherwise, pockets of fuel will not interact with the oxygen in the exhaust gas stream, and the result will be incomplete combustion (i.e. combustible emissions). However, with TEG it is equally important to avoid quenching the combustion process with too much localized mixing.   This mixing process is accomplished through the design of the duct burner. The Coen PowerPlus® low emission duct burner has been designed to provide superior mixing of fuel and GT exhaust gas without quenching.

The third factor for complete combustion is time.  Although a combustion process may have adequate temperature and adequate oxygen (turbulence), the process requires time to complete.  Even though this time may be only a matter of milliseconds, in the highly chaotic region directly downstream of the duct burners, where the flames are visible and the reactants and products of combustion are all intermixed, some reactants exit this “reaction zone” before the combustion process can be completed.  Therefore, the term “residence time” is typically used in the industry to define the amount of time reactants “reside” in the reaction zone.  An increase in residence time typically results in a decrease in combustible emissions, provided all the other factors are in compliance as well.

It is important to remember that thermal oxidation of TEG combustible emissions takes place in every duct burner system to some small degree.  TEG is the only source of oxygen for the burner fuel, so some of the TEG combustible emissions are always destroyed in the burner flame.  Typically, any benefits derived from this oxidation are more than offset by the combustible emissions generated by the duct burner flame.  Therefore, to achieve any benefits from thermal oxidation, it is important that the burner   generate as little combustible emissions as possible.

COMBUSTION EXPERIENCE AND ANALYSIS:

Utilizing field emission data from PowerPlus® duct burner applications as a base line and projecting system performance using CFD modeling along with other proprietary models, the study results indicate it is feasible to thermally oxidize 50% of the combustible emissions generated by the gas turbine.  However, changes in both the duct burner and HRSG designs, as well as their operational requirements, would be needed to achieve this optimized level of thermal oxidation.

For several applications that are currently in operation, emissions data has been collected by an independent testing laboratory and analyzed.  Table 2 illustrates some of the emissions that were noted in the field under controlled operating conditions.

Table 2

(PowerPlus® Burner , 18 foot furnace, 1300°F Fired Temp.)

Emission Rates

CO

lbs/hr

Unit #3

CO

lbs/hr

Unit #4

From GT

2.7

2.2

After Duct Burner

2.5

2.0

Reduction – %

7.4

9.1


The thermal oxidation process, to some small degree, has been demonstrated in these actual applications, as shown above.  However, further analysis indicates it is possible to achieve an even greater percentage of thermal oxidation.

To complete this analysis, the first step was to create a model of the combustion system in the HRSG.

Figure 5 below illustrates the duct burner configuration used for the CFD modeling.

Figure 5, Burner Model Configuration

To achieve the higher levels of thermal oxidation, up to 50%, a cooperative design effort would be required between the HRSG OEM and the duct burner manufacturer to develop the desired environment to produce efficient oxidation of residual combustibles in the TEG and from the burner.   The design features considered the most critical for the system are:


  • A low emission duct burner design
  • Retaining a bulk mixed exhaust gas temperature of 1300 °F or greater
  • Uniform TEG flow distribution over the duct cross section before the burner
  • Use of baffling to direct the flow of TEG near the burner flames
  • Increased furnace residence time

We have chosen a furnace firing depth of 28 feet as a reasonable design for this analysis. The HRSG furnace length supplied on a typical   “F” class GT application varies with the HRSG manufacturer from 12 feet to 18 feet in length.  The selected furnace length represents an addition of 86% in furnace volume compared to the average OEM design.

Figure 6. below illustrates the TEG flow velocity and distribution through the burner in the CFD model.




Figure 6, CFD Model

The bulk mixed exhaust gas temperature of 1300 °F is considered the lowest acceptable temperature to achieve the target goal of a 50% reduction of combustibles in the TEG with a 28-foot furnace.  At temperatures below 1300 °F, the thermal oxidation effect still exists, but the percentage reduction of the TEG combustibles will be diminished.

For the model, CO and CH4 (methane) were used as the basis for predicting the oxidation of combustibles in the exhaust gases down stream of the burner.  As denoted in Table 1, CH4 and CO require higher temperatures than the other combustible species for thermal oxidation.  This simplifies the model as it allows CH4 to represent all Unburnt Hydrocarbon (UBHC) without degrading the value of the model results.

The modeling results for the maximum (Case 1 @ 850 MM BTUH, 100% GT load, 750 MW) and minimum (Case 2 @ 225 MM BTUH, 50% GT load, 370 MW) fired operating conditions for the NovelEdge system are shown in Figure 7.  This figure is split with two views, a full range graph, and a close up view of the lower range of the full graph at the extended furnace lengths.  This graph shows the GT emissions for each case upstream of the burner and then at 4-foot intervals down the length of the furnace.  (Note: The model shows the extrapolated results for a 32-foot furnace length.)

Figure 7, Oxidation Charts:


A longer furnace length is required to achieve the oxidation of CH4 to a level equal to 50% of it’s inlet mass than is required to achieve CO oxidation to 50% of it’s inlet mass level.  Coen expects that near equal results can be achieved using a shorter furnace length with more refinement of the model and with the correlation of additional test data.


OXIDATION CATALYST:

Obviously, it can be argued that an oxidation catalyst can also be utilized to reduce some combustible emissions.  This is true.  However, there are drawbacks to utilizing an oxidation catalyst, and they are as follows:


  1. It will add approximately $1 million to each HRSG for an “F” class GT
  2. It will add back pressure and lower GT output and efficiency
  3. The life expectancy for the catalyst is a nominal 4 years
  4. Outside of its temperature range, the catalyst effectiveness is greatly reduced

Given these reasons, it is not surprising that only a modest percentage of the “F” class GT combined cycle installations in recent years have included an oxidation catalyst.

The other factor for the oxidation catalyst is its effectiveness.  For CO, at its optimum temperature, the oxidation catalyst is 80 to 90% effective.  For VOC’s, again in its optimum temperature range, the oxidation catalyst is 30 to 40% effective.  For carbon particulates its effectiveness is 0%.

Secondly, there is no reason an oxidation catalyst cannot be included with this new technology plant.  Therefore, if one was to calculate the math, the rate of reduction in combustible emissions is still a factor of 3 to 1 as illustrated previously, as long as both plants include the oxidation catalyst (the new technology HRSG would include the oxidation catalyst downstream of the duct burners).  And as a matter of discussion, the catalyst for the newly patented combined cycle facility will have less duty than the conventional plant, so it will have a less expensive catalyst.  In addition, a 750 MW new technology application only requires 2 catalysts in lieu of the 3 required in the conventional cycle.

In summary, in some instances, the low-emissions technology used in the NovelEdgeTM cycle can eliminate the need for a CO catalyst, as the duct burners will incinerate emissions to a comparable level, or in some instances, such as that of particulates, reduce emissions that are unaffected by the oxidation catalyst.  In applications where very low emissions are required, the new technology facility can include an oxidation catalyst, which will further reduce the CO and VOC’s, which have already been reduced to approximately 1/3 of the conventional plant emission levels in the PowerPlus® duct burners.  This results in a reduction to less than 7% and 25% of untreated levels for CO and VOC’s respectively.

CONSIDERATIONS FOR NOx:

The technology discussed herein has the ability to attain large reductions in combustible emissions, namely CO, UBHC, VOC’s, and combustible particulates.  However, this still leaves the question regarding one of the pollutants of greatest concern, NOx.

NOx is produced in both the GT and in the duct burners.  Per Btu of heat input, the duct burners produce significantly more NOx than a GT which includes a Low NOx burner system, in some instances as much as 2.5 times as much.  Therefore, the total amount of NOx formed in the new technology cycle is greater than that of the conventional unfired cycle of equal output.

However, when a Selective Catalytic Reduction (SCR) system is employed, the advantage can be shifted in favor of the new technology.  Much of this advantage is driven by economics.  In the conventional unfired system, the NOx level into the SCR catalyst contains only NOx produced by the GT.  As this level of NOx is reduced in the SCR, it reaches a concentration whereby it becomes increasing difficult (expensive) to reduce further.  Thus, for a conventional unfired application, this SCR may have an efficiency of 78% and an outlet concentration of 2.4 ppm uncorrected, or 2.0 ppm corrected to 15% oxygen.

For the application utilizing the new technology, NOx is produced both in the GT and in the duct burners.  The level of raw NOx into the SCR is typically 2.0 times the mass flow of NOx into the SCR in the conventional unfired cycle (on an HRSG to HRSG comparison – does not include the quantity 2 versus quantity 3 HRSG factor).  Again, this higher level of NOx is reduced in the SCR, until it reaches a concentration whereby it becomes increasing difficult to reduce further.  Thus, for this new technology application, the SCR may have an efficiency of 91% and an outlet concentration of 2.4 ppm uncorrected, or the same outlet concentration as in the conventional design.  Since the oxygen has been consumed in the duct burning process, the new technology outlet NOx, when corrected to 15% oxygen, is 1.25 ppm.  This yields an equal outlet stack mass flow of NOx from both the conventional and the new technology application.

For the new technology application at 91% efficient versus 78% efficient, more catalyst is required in each new technology SCR system.  This adds cost to the individual SCR systems.  However, since the 750 MW new technology combined cycle requires only 2 SCR systems, versus 3 SCR systems for the conventional plant, the total cost for SCR systems for either plant is similar.  A major supplier of SCR equipment has verified this phenomenon.

Therefore, for an equivalent investment in up front capital, and with equal mass flow of NOx from each HRSG stack, and given only 2 HRSGs versus 3, the NovelEdgeTM cycle can benefit from a one third reduction in NOx, in addition to the large reductions in combustible emissions.

A 750 MW EMISSIONS COMPARISON – CONVENTIONAL 3×1 VS NovelEdgeTM 2×1:

Utilizing the CFD analysis, factory test results, and field test data, it is anticipated that the incineration rate for this technology will be on the order of 50% for the combustible emissions.  This 50% rate of thermal oxidation, however, although predicted through CFD modeling and analysis, has not yet been proven in a real world application.  Therefore, the first application will only guarantee a low rate of burner emissions, due to commercial considerations.

After the first unit is tested and the data is analyzed and correlated to the CFD model predictions, lower emission levels can be guaranteed.  This process will evolve with subsequent units, until all equipment manufacturers can be comfortable with the emissions guarantee and the degree of margin that is associated with that guarantee.  The long-range goal is to be able to guarantee a thermal oxidation rate that approaches 50%.

Table 2 provides a comparison of the emissions from an unfired 750 MW conventional plant in a 3×1 arrangement with “F” class GT engines, to a 750 MW NovelEdgeTM plant with the same model GT engines (only 2 instead of 3).  The NovelEdgeTM Guarantee reflects the numbers associated with the first application for this technology.  The NovelEdgeTM Expected data reflects the numbers consistent with a 50% thermal oxidation rate. As can be seen from the data in this table, the combustible emissions are reduced nominally by a factor of 3 for this technology.

Table 3

Emissions Rates1 (Tons/year)4

NOx

CO

VOC

PM10

Carbon Particulate

Conventional Plant

202

378

37

118

18.0*

NovelEdge Guarantee2

208

420

50

128

26.7**

NovelEdge Expected3

2085

125

12

108


6.1

*  Assumed 15% by weight of total particulate from CT.

** Assumes 30% by weight of the total particulate from the burner.


  1. Emissions rates are during steady state operation at base load.  Both configurations include an SCR system to reduce NOx emissions to 2.5 ppmvd @ 15% O2.
  2. Based on the combustion turbine and duct burner manufacturers guaranteed emissions values.
  3. Expected emissions with 50% incineration of total combustibles.
  4. Based upon potential to emit, full load, ISO conditions, and 8760 hours per year.
  5. With total SCR costs similar to the total conventional 3×1 plant, this number could be reduced by up to 33%.
SUMMARY:

The NovelEdgeTM combined cycle technology, in conjunction with a PowerPlus® duct burner system, and other enhancements provided in the HRSG by Nooter/Eriksen, can provide a low-cost, efficient, ultra-low emission power plant.  The overwhelming advantages of this system include:


  1. Huge reductions in combustible emissions are possible, a factor of 3 reduction is the targeted goal.
  2. The system is simple; there is no need for a catalyst, reducing agents, chemical storage systems, etc.
  3. The NovelEdgeTM plant costs less to own and operate than the conventional combined cycle power plant.
  4. Future applications are projected to have the added ability to reduce emissions related to start-up and shutdown operations

With this new technology, the Team of NovelEdge Technologies, Coen, and Nooter/Eriksen believes they can offer this ultra-low emissions technology to reduce emissions, yet at the same time provide a lower cost facility than the conventional combined cycle power plant.  Lower emissions for less money, truly a win-win situation for everyone.




1 See PowerGen 2002 presentation entitled “A Cost Effective Combined Cycle for the 21st Century”, by Edward DePietro.