The natural gas transmission industry has been subject to increasingly stringent emissions regulations over the years, with more to come. One big challenge has been to reduce oxides of nitrogen (NOx) while maintaining existing engine reliability and power output. One way to lower emissions from reciprocating engines is to use technologies that increase engine air flow. To do this, the engine can be turbocharged, or, when a turbocharger already exists, the turbocharger is “upgraded” to increase air flow rate. However, the high engine air-flow rates required for an engine upgrade project are constrained by the physics of the turbocharger. As air flow through
an engine increases, exhaust temperature decreases accordingly. While lower exhaust gas temperature reduces NOx emissions, it also reduces available energy needed to drive the turbocharger turbine.

A second challenge with an upgraded, turbocharged engine system faces is maintaining mass flow through the compressor as ambient conditions change. As ambient temperature of the compressor inlet air increases, density of this air decreases, which results in a lower mass flow rate to the engine. To maintain the same mass flow rate, the turbocharger must rotate at a higher speed, requiring more power from the turbocharger turbine. Often times, there is not enough energy available for the turbocharger turbine to rotate at the required higher speed.

To overcome these problems, a turbocharger booster system using a relatively small dry low emissions (DLE) burner was tested in a laboratory. The prototype proved that a burner placed in front of the turbocharger nozzle ring could raise turbine inlet temperature, increasing the turbocharger rotating speed.

Despite significant engineering done on the laboratory prototype, several challenges had to be addressed to make this system field installable. For industry acceptance, the field booster system needed to be a cost-effective alternative to a new turbocharger, engine replacement, or other boosting methods — without impacting overall emissions. Finally, the system must be practical to install and maintain.

A new system derived from the laboratory approach was installed on a field engine. By using the burner to add energy to the turbine, turbocharger speed and air mass flow rate could be increased. This created the capability to overcome situations that would otherwise require the engine to run at a lower horsepower or necessitate a costly turbocharger replacement. By using a small amount of auxiliary air from the compressor discharge, appreciably increased emissions were avoided. The field-installable turbocharger booster system was successfully developed. It was proved that a small, efficient burner used to independently control the available energy at the turbocharger turbine inlet could increase the operating range of an actual engine without detrimentally impacting the existing engine or turbocharger.

First steps taken

The National Gas Machinery Laboratory (NGML) serves the natural gas transmission industry by developing and researching emissions reduction strategies for prime movers used in pipeline service. One key component to emissions reductions on the reciprocating engines used
in pipeline service is air-flow delivery.
On two-stroke cycle, lean-burn engines, air delivery to the engine typically is accomplished using a scavenging pump, fan/blower, or turbocharger. The intention here is to concentrate on engines with turbochargers as the prime source of air delivery.

In a lean-burn, two-stroke cycle engine, the production of NOx is directly related to in-cylinder peak temperatures during combustion (Heywood and Sher, 1999). To decrease NOx production rate, a lower combustion temperature is required. Lower combustion temperatures can be achieved with an increased trapped air-to-fuel ratio attained by increasing the amount of air provided by the turbocharger. The turbocharger needs to sustain the required air flow for the benefit of lower NOx over a wide range of operating conditions.

Air-flow rate from the turbocharger compressor is influenced by several thermodynamic constraints. Figure 1 illustrates a turbocharged engine configuration as it operates in the field. Air is drawn in by the compressor inlet (#1, Figure 1), compressed, and delivered to the engine from the compressor discharge (#2, Figure 1). The engine goes through its combustion cycle and the exhaust gases are sent to the turbine inlet (#3, Figure 1) where the exhaust energy is consumed to power the turbocharger before exiting the system at the turbine discharge (#4, Figure 1). Power developed in the turbocharger turbine precisely balances with the power required by the turbocharger compressor to deliver the necessary amount of air flow.

To achieve the desired lower combustion temperature, and hence lower NOx production, the trapped air-to-fuel ratio in the engine must be precisely controlled and oftentimes increased. The turbocharger compressor must, then, deliver a greater amount of air at an increased pressure ratio, requiring a greater amount of power from the turbocharger turbine. However, the lower combustion temperature leads to decreased available exhaust energy, which ultimately means there is less power available at the turbine inlet to power the compressor.

A second challenge the upgraded turbocharged engine system faces is maintaining mass flow through the compressor as ambient conditions change. As the ambient compressor inlet air temperature increases, density of this air decreases. The volumetric flow rate through a compressor stays relatively constant despite these density changes; however the mass flow rate, which ultimately controls combustion temperature, can be significantly reduced by higher ambient temperatures. To maintain a mass flow rate that is independent of ambient temperature, the turbocharger must rotate at a higher speed, which requires more power from the turbocharger turbine.

In both these conditions, if the
turbocharger is designed correctly, excess energy will be available at the
turbocharger turbine. The excess energy is diverted away from the turbine by a waste gate, allowing the proper power balance between turbine and compressor. Nonetheless, even with the waste gate fully closed, the required compressor power could exceed the available turbine power, if the turbocharger was improperly designed, has suffered performance degradation over time, or has new operating conditions imposed. These shortcomings led to recognition of the need for a system that can “boost” the performance of the turbocharger during times of inadequate power.

Peaking system suggested

The turbocharger drawbacks previously described, along with the need to improve NOx reduction strategies, led to recognition of a need to improve on current
turbocharger technology. Therefore, the Pipeline Research Council International, Inc. (PRCI) Pipeline Engine Emissions Control Roadmap (2005) suggested that a turbocharger peaking system be used to solve this operating challenge. The peaking system must increase turbocharger and engine performance by providing constant air mass flow rate to the engine regardless of ambient conditions, without increasing overall emissions.

Several different methods are used
to remedy concerns with current turbocharger technology. The NGML has been conducting research on a variable geometry turbocharger, which allows the diffuser vanes to be rotated from the OEM position in real time. By changing the vane angles, turbocharger compressor performance can be optimized for any reasonable operating condition. Setting the vanes at different positions depending on ambient temperature alters the turbocharger for hot or cold seasons.

A second method uses an afterburner. The afterburner adds a small amount of fuel at the exit of the gas turbine engine, which then combusts and increases exhaust temperature, thereby increasing available energy at the power turbine inlet. To apply this concept to a turbocharger, a small natural gas burner could be added between the engine exhaust manifold and the turbocharger turbine inlet. Combusting a small amount of fuel increases the exhaust gas temperature, effectively increasing the thermal energy in the exhaust stream. Increased available thermal energy allows the turbine to produce additional power.

A third option would be to cool the engine intake air using a chiller. This would increase the density of the intake air and raise the mass flow rate.

A final option uses an external compressor to achieve higher pressure and flow rate of engine intake air. Again, this option would increase density and boost the mass flow rate of air delivered to the engine.

After investigating the available options, a small natural gas burner was selected because of its lower annual operating cost. The next step was to design a turbocharger booster system that uses a small natural gas burner, and perform baseline testing at the NGML Turbocharger Testing and Research Facility. This phase was completed in the PRCI project: Development of a Turbocharger Booster System.

The first step was to conduct a thermodynamic analysis of exhaust gases entering the turbocharger turbine to determine the amount of energy needed to increase exhaust gas temperatures to sufficient levels. Another concern with using a natural gas burner is that increasing exhaust stream temperature might also increase NOx and CO overall emissions levels. The researchers determined that a lean combustion burner would reduce NOx but might produce unwanted CO. Another issue researchers discovered was despite relatively high oxygen content in the exhaust stream of large two-stroke cycle engines, oxygen content was not high enough to sustain the desired stable lean-burn combustion.

Given the operating and emission constraints, the research team selected the suitable burner and an auxiliary air stream, illustrated in Figure 2, to meet the requirements constraints.

Without the turbocharger booster system (TuBS), this enthalpy flow rate equals the enthalpy flow rate from the engine. The goal of TuBS is to increase the enthalpy flow rate by increasing the turbine inlet temperature and mass flow rate to the turbocharger turbine.

The isentropic outlet enthalpies are functions of the inlet conditions and
the pressure ratio across the particular machine. The actual power consumed
by the compressor must equal the actual power produced by the turbine multiplied by the turbocharger efficiency. An increase in hin,t increases the power produced by the turbine. Because the turbocharger must operate in an equilibrium condition, turbocharger speed increases until the power consumed by the compressor equals the power produced by the turbine less turbocharger inefficiencies. The result is increased mass air flow at a higher pressure through the turbocharger compressor and into the engine.

As discussed above, changes in
ambient temperature deleteriously affect the mass air flow rate through the turbocharger compressor. An increase in temperature is coupled by a corresponding decrease in gas (air) density. Many turbocharged engines operate in environments with temperature ranges from below the freezing point in the winter to above 100°F in summer. It becomes clear that the density of air is greatly affected by these temperature changes. While a turbocharger compressor delivers a constant volumetric flow rate at a given speed regardless of temperature, the mass flow rate will change with the density.

The air-to-fuel ratio, depends on the mass flow rate of air and fuel, not the volumetric flow rates. This leads to the other drawback discussed — which is the exhaust-temperature trade-off inherent to lean-burn turbocharged engines. A higher air-to-fuel ratio lowers exhaust temperature and NOx emissions, while at the same time decreasing available power to the turbocharger turbine. This is also of consequence on a day with high ambient temperatures; the engine will see a decreased air mass delivered to the engine with a constant mass of fuel.

The need, then, to independently control the turbocharger is clear. Increasing ambient temperature and/or increasing air-mass flow rate through the engine creates a deleterious effect on turbocharger performance — an effect
that can be mitigated by independent turbocharger control.

Laboratory prototype built

The design of the turbocharger booster system (TuBS) used in the laboratory test consisted of a burner, a gas train for fuel delivery, provisions to bleed a small amount of air from the compressor discharge piping for the burner, and a control system. The Kinemax G-series burner was selected because it did not install in-line with the exhaust stream, but rather flange-mounted externally to the exhaust stream. This results in a longer life burner because it is not subject to exhaust gas stream conditions. The Kinemax burner also operates at lean air-to-fuel ratios which help produce minimal levels of NOx.

A natural gas train was designed to safely and reliably deliver fuel to the burner. The gas train consists of three automatic block valves, each of which is independent and fails “closed” to ensure safe operation. A pressure switch trips the block valves closed, if pressure in the combustion chamber rises above that of the supply line, thus preventing the flame from traveling back into the supply line. Finally, the gas train includes a fuel control valve, coupled to an electro-pneumatic actuator, to control the amount of fuel supplied to the burner.

A valve to control the air, in addition to the fuel valve on the gas train, is used for accurate control of the air-to-fuel ratio at which the burner operates requires. Both these valves are controlled with an analog output current signal from an OPTO 22 control system. To determine the air-to-fuel ratio, both air-flow and fuel-flow rates were measured using an orifice. In the laboratory prototype, the valves could be modulated independently of each other to vary the burner air-to-fuel ratio.

Several challenges had to be addressed to make this system installable in the field. The system also had to remain cost effective compared to a new turbocharger, engine replacement, or other boosting methods. Additionally, overall emissions from the engine and booster system had to be constant or reduced. Finally, the system had to be practical to install and maintain in a safe manner.

To move the prototype from the laboratory to the field, the NGML partnered with Exline, Inc. in Salina, Kansas. Exline is a highly regarded company with over 100 years of experience serving the natural gas transmission industry.

The next field-testing step was to select a suitable site. The project team selected engine 515 at the Panhandle Energy Haven, Kansas compressor station. This Clark TLA-10 two-stroke cycle was turbocharged with an off-mounted turbocharger that was located outside as shown in Figures 3 and 4.

Components were modeled and sized for this specific engine. The size of the burner is a very critical match. If it is too small it will not be able to provide enough energy to fully overcome degraded turbocharger performance. If it is too large, it will operate inefficiently, with a potential impact on emissions from the burner. Figure 5 illustrates the method of installation on the engine with auxiliary air extracted from the compressor discharge after the intercooler.

The TuBS burner increases mass flow rate at the turbocharger turbine inlet, along with increasing the turbine inlet temperature.

As previously described, overall system energy is conserved, so compressor and turbine work must equal each other. Because the TuBS adds energy to the turbine, a second energy balance had to determine how the required turbine inlet temperature is related to compressor inlet temperature, which is the same as the ambient air temperature.

As ambient temperature increases, to maintain all other factors, turbine inlet temperature must increase. To determine the exact amount of temperature increase, a combination of actual engine data and turbocharger test stand data was required. Actual engine data was provided by Panhandle station personnel, which allowed specification of parameters for testing the turbocharger. Using those parameters, the turbocharger was load line tested at the NGML TTRF. A load line test simulates the exact restriction the engine and cooler places on the compressor. The load line data provides two key pieces of information: 1) turbocharger efficiencies to enable the use of equation (11) to determine actual required turbine inlet temperature and 2) the means to calibrate the controller for the TuBS system, the Digital Turbocharger Monitoring System (DTuMS), soon to be discussed.

With turbocharger data, calculations would require turbine inlet temperature for the worst-case expected ambient temperature. That turbine inlet temperature can then be compared with normal engine exhaust temperatures to determine energy input required by the TuBS.

For this Clark TLA-10 engine,
modeling analysis determined that 1.4 MMBTU/Hr was required from the TuBS burner. The Maxon Kinemax burner was a 4 MMBTU/Hr burner — too large for this application. After investigating available options with the same criteria as required for the laboratory prototype, an Eclipse ThermJet burner was selected. This burner has several key advantages that make it ideal for field installations. Integrated fuel and air orifices for flow rate measurement as well as field changeable inlet clocking will simplify future installations, plus these burners are available in a wide range of sizes from 0.150x106 BTU/Hr to 20x106 BTU/Hr.

With burner sizing complete, the air and fuel delivery system was designed. At the Haven compressor station, all the engine fuel runs were on the opposite side of the building from the turbocharger. This meant a gas line had to be run the width of the engine room to the gas train near the turbocharger. Like the one installed for the laboratory prototype, this gas train, was designed to safely and reliably deliver fuel to the TuBS burner with failsafe operation. Fuel was first regulated down from 80 psig to approximately 15 psig by a natural gas regulator, and then passed through an orifice plate for flow measurement Lastly a V-Ball valve was installed for gas flow control and shut off operation. A secondary block-and-bleed valve was also installed at the fuel connection point for the engine controller. The fuel train was instrumented for pressure and temperature, allowing the controller to monitor for fuel delivery status.

Due to the layout of the turbocharger skid, auxiliary air from the compressor discharge could only be extracted after the intercooler. Depending on the specific installation, auxiliary air could be extracted before the intercooler. The air train was very similar to the gas train, with an orifice for flow measurement and a V-Ball valve for flow control. A regulator was not necessary because the control valve was sized to compensate for small changes in pressure.

Fuel and air were piped to the burner, installed in a burner tube connected
to the engine exhaust before the turbocharger turbine. The burner tube was meant to contain the entire flame to alleviate any chance of flame impingement against the opposite wall of the exhaust manifold. An external burner tube would also protect the flame from aerodynamic instabilities associated with direct exposure to exhaust flow. Additionally, the burner tube could be standardized, which allowed any installation requiring this burner size to use the same design with the only change being the transition from the burner tube to the exhaust manifold connection. This modular design concept is key for cost- effective and practical field installations.

This engine had an existing six-inch flange on the exhaust manifold that was used as the connection point for the burner tube. In this field installation,
the burner tube was installed facing
down from the exhaust manifold.

A method for simple, robust and cost effective TuBS control was required. The parameters of interest for the testing of the booster system were: the turbine inlet temperature, the compressor air mass flow rate, and the pressure ratio across the compressor. After searching for commercially available options, ScavengeTech, LLC was contracted to provide a DTuMS. The DTuMS is a state-of-the-art monitoring system comprised of a compact sensor package with turbocharger analysis and trending software. Measured parameters include
turbine pressure and temperature, compressor pressure and temperature, turbocharger speed, compressor mass flow, and optional vibration measurement. This sophisticated turbocharger sensor package was an off-the-shelf option for all the critical information needed for operation and control of the TuBS. ScavengeTech designed additional inputs and outputs to allow for interfacing and control of the TuBS components. This entire sensor and control package is available for future field installations.

Field prototype testing

To test the turbocharger booster system on a field engine, a test plan was followed to make sure all impacting parameters were tested and results complete. The first objective was to assess impact of the burner air-to-fuel ratio on overall emissions from the engine burner system. By determining the burner air-to-fuel ratio that produced the least amount of emissions, all ensuing test points could be taken at that operating condition. The varying burner air-to-fuel ratio also permitted a complete assessment of the burner robustness. The second objective was to characterize interaction of the booster system with the engine and how that could be applied to normal operation. Data included turbocharger parameters from the TuMS system, engine performance information, and emissions data after the turbocharger turbine. The emissions data was collected using an ECOM J2K-N portable analyzer.

The test plan consisted of a step-by-step listing of all test operations, including initial safety compliance and operational functionality, baseline engine data, and steps to begin testing the turbocharger booster system. The initial baseline data consisted of engine runs at 100% speed with torque ranging from 92% to 98%. An initial baseline was also completed with reduced engine ignition timing because this is a common method of offsetting the effects of reduced turbocharger performance. After baseline data points were complete, the ignition point of TuBS was found, which represents the minimum energy input possible from this burner. From this point on, the gas flow was increased to allow for testing the TuBS at different energy inputs while also testing different air-to-fuel ratios in the burner. To verify the testing results were repeatable and accurate, the system was allowed to reach thermodynamic equilibrium between data points.

A summary of collected data at 98% load was collected for comparison purposes. These points characterize the booster system and validate that the burner was able to increase turbine power, which in turn increased compressor mass flow rate, pressure ratio, and speed, if there was no wastegate margin. Impacts seen on the turbocharger were a significant change in turbine inlet temperature and wastegate margin. The concept was verified by starting the booster system and observing the increased power delivered to the turbine side.

Figure 6 shows the effects of the TuBS energy input 1 on the turbocharger and Figure 7 shows the same information except for energy input 3. On both charts all parameters except the turbine inlet temperature use the Wastegate Position scale as described in each legend. The chart in Figure 6 shows that even though the turbine inlet temperature increased by an immeasurable amount, a significant increase in the wastegate margin was realized. Comparing Figures 6 and 7 also illustrates an increase in wastegate margin. However, the graph in Figure 7 clearly shows the increase in turbine inlet temperature, which is another direct indication of increased available power at the turbine inlet. By using the airflow load line from the turbocharger test, preliminary analysis data indicates the TuBS energy increase corresponds to an approximate 10% increase in available air flow rate. n

Acknowledgements

The authors express thanks to the Pipeline Research Council International, Inc. (PRCI) for supporting this project and making the idea of a turbocharger booster system a reality. The authors also thank Panhandle Energy and the staff at the Haven, Kansas location for their assistance in testing this prototype on their engine. The authors are grateful to Exline for providing the installation and the many suggestions to improve the system and to Exterran for providing a TLA-10 turbocharger case to assist testing the turbocharger at the NGML. Finally, the NGML thanks ECOM for donating the portable emissions analyzer at the 2007 GMC that was used to measure the CO and NOx emissions during the field test. Based on a paper presented at the Gas Machinery Conference held in Albuquerque, New Mexico, October 6-8, 2008.