Thermoelectric generator for engine control system

Final Report Summary - THETAGEN (Thermoelectric generator for engine control system)

Executive Summary:
ThetaGen aims to scavenge thermal energy from a helicopter turboshaft and convert it into electrical power to feed the engine control unit.
It defines and specifies such a power block, benchmarks potential technologies, develops a demonstrator and proves the concept thru actual measurements on an engine.
It gathers companies over four European countries: an aircraft manufacturer (EVEKTOR, Czech republic) , a research lab (CNRS, France), an expert in thermoelectric devices (TERMOGEN AB, Sweden), a thermal expert (EPSILON, France) and a power block manufacturer (TRONICO, France).

Project Context and Objectives:
1 Progress beyond the state of the art
Thermoelectricity is not a new matter and a strong scientific effort has been made in recent years. Important progress has been made and the key technology points are now well understood, even if a lot is still to be done.
Right now, it is clear that this technology can offer many interesting solutions in energy scavenging.
It is time to push the scientific results into actual applications and develop added value solutions.
Currently, only a few products using thermoelectric phenomena are on the market. Most of them are gadgets or toys more than real products usable for practical technical applications. Several items are just emerging for remote sensor applications, mainly in the domestic field.
Specifically, these techniques have never been used in aircraft systems. Thus, a lot of heat generated by aircrafts is currently wasted around the engines and coolers and it could be usefully harvested.
Applications of this extra energy in an aircraft are endless, but our project focuses on the engine itself. We want to generate enough energy to feed the electric devices controlling the engine. In that way, the engine control unit can be fed in autonomy from the heat, independently from aircraft electric power system.
The main advantage of this thermoelectric generator is to provide an inherent failure recovery mode if the main electric system should fail. In that case, the engine stays under control of the FADEC and all parameters can still be logged in memory for further analysis.
In addition, a thermoelectric generator directly increases safety of flight in power system failure mode, because more emergency capacity of batteries remains available for avionic systems and further it does not decrease engine performances like FADEC generators.
To obtain this gain, many latches have to be opened:
- Specify clearly the requirement for such a engine thermo generator;
- Map the thermoelectricity landscape and chose the best architecture to transform heat into electricity;
- Disclose and solve thermal problems;
- Achieve a robust thermo generator design suitable for harsh environment operation;
- Overcome power conversion issues;
- Address safety, reliability and airworthiness issues;
- Prove the concept and demonstrate the value of such a generator.
According to our expertise, the current state of the art in the thermoelectric field allows building valuable system based on the thermoelectric phenomena, nevertheless such an approach has never been attempted.
In addition to the project targets, it is clear that we will learn a lot about thermoelectricity in avionic systems, yielding new ideas and opportunities. This project is pioneering this field and is going to open the way for new applications, and not only in the aeronautical world.
To address these problems, the consortium has to bring together several different experts, not used to working together:
- An aircraft manufacturer, expert in electrical and mechanical problems in electric systems and engine installations of aircrafts;
- A thermo generation expert, with skills from former applications and products;
- A thermal expert for thermal architecture definition;
- An electronic expert for power conversion and power coupling;
- An academic scientist for scientific matters, benchmarking and bibliography.

2 Scientific and technology methodology and work plan
The work plan is divided into four work packages plus one WP for management (WP0).
The precisely defined management approach in ThetaGen, the high qualification and experience of the management team gained in previous EU Framework Programmes as well as the use of the modern communication and management tools are the good starting point for the ThetaGen management structure.
The consortium management of the ThetaGen project is implemented through the WP0. The management structure aims at ensuring timely and qualitative achievement of the results, and at providing timely and efficient legal, contractual, financial, and administrative coordination of the project.
Individual Work packages strictly respect the workflow required in the call text.

WP1 Statement of Work
The crucial point of this work package is a very close cooperation between the consortium and CLEANSKY when all needs and constraints of the demonstrator are collected, analysed and fixed in the commonly agreed basic document (Statement of Work) comprising high level input specifications:
- Electrical parameters and interfaces of the engine control unit (FADEC) in all operational phases and the range of engine management;
- Mechanical and thermal interface with the turbo shaft;
- The constraints from the environment: mechanical, thermal and electrical constraints, RTCA-DO160 levels of compliance.
This SoW is a deliverable.

WP2 Feasibility Proof
The state of the art of thermoelectric generation is studied in detail in various industrial applications. Works focus on the latest experiments and demonstrations worldwide so as to provide benchmarks for our system.
This benchmark is synthesised in a deliverable document, showing clear conclusions about demonstrator feasibility.

WP3 Technology Development
This work package includes the entire task to develop and implement a product from a clear specification:
- Overall architecture definition
- Detail definition of sub modules
- Sub modules design, fabrication and testing
- System integration
- Complete system trials in our labs
- All relevant documentation

WP4 Integration and experiments
The prototype developed is delivered to a turbo shaft manufacturer for engine integration.
This manufacturer is responsible of all the tests and trials.
The consortium supports the integration on the test bench and trial campaign.

Project Results:
1 CONCLUSIONS
The project failed to design the expected demonstrator because:
1. A current turboshaft is not a friendly environment for a thermo generator. It has not been designed for, and it does not fit with any evolution. No practical solution could be found with the current design;
2. In future turbo shafts, several evolutions have to be considered: anchoring devices on the combustion chamber and/or the compressors, cooling the fuel circuit and/or the fuel tank, improve the oil circuit…
3. A wide range of temperature is definitely not practical at system level. To narrow the temperature domain in all flight conditions it should be nice to use the combustion chamber as a hot source (more or less stable) and the oil circuit as a cool source (temperature controlled for other reasons). The fuel circuit can also be a nice cool source if properly refreshed.
In any solution, a global system analysis at the engine level has to be held.
2 TECHNICAL CONSIDERATIONS
2.1 System aspects
A thermoelectric system is a complex arrangement able to deliver power in an optimal way, only in very specific conditions. This has been detailed in WP2_D01, but let us summarize in the following way:
1. DC power delivery from the DC/DC converter is dependant of the thermo cells internal resistance, itself dependant of the temperatures. It has been demonstrated that the maximum power can be delivered when the internal resistance of the cell arrangement equals the input resistance of the DC/DC converter. Depending of the flight conditions, the temperatures change over a wide span, and optimum power transfer is not natural. The DC/DC converter shall embed a specific circuit able to determine the best input resistance and to implement it. Such a system is very well known (MPPT) but not straightforward to implement.
As a conclusion, we can say that the optimum electrical power transfer can be implemented with some electrical complexity.
2. Thermal transfer from the sources to the thermo cells is also dependant of the thermal resistances. It has been demonstrated that the maximum efficiency happens when the internal thermal resistance of the cell arrangement equals the total thermal resistance of the heat exchangers. Unfortunately, this maximum efficiency point is very different from the maximum output power delivered. We have to make a choice between power and efficiency, but at the end of the day, for a given output power, the best possible efficiency is very poor. It means that we need a lot of thermal power, more or less incompatible with other constraints: mass, size, transfer coefficients…
The best possible efficiency is not practical, as it depends of the cells temperatures. It can be achievable for a given working point of the turboshaft, but not for the wide span of operating point. As far as we know, a technique to dynamically adapt the operating point to the inputs does not exist. Practically, it means that the system has to be massively oversized to ensure power delivery in every condition. This oversizing has not been possible in the given requirement.

2.2 Hot source considerations
A standard turboshaft delivers 1 MW of mechanical power with an average efficiency of 30%. So basically, it wastes 2 MW of thermal power. Thus it should be possible to scavenge enough thermal power to feed our thermo cells. Unfortunately, things are much more complex than that.
On the turboshaft we have 3 options to catch heath:
1. On the compressor segment: compression of air produces a lot of heat. This is quite positive as this heath contributes to the mechanical power generation after the combustion chamber. We have there available a lot of heath close to the intake of the turboshaft. A very nice solution could be to use this compression segment as a hot source and the air intake as a cold source. In this way we have a very short thermal path and we don’t waste useful heath as all the power caught from the compression segment is re-injected in the air intake to be compressed.
We can also use anything else than air to cool the thermo cells, but we waste the scavenged heath. This is not really an issue as it represents more than a few tens of per cents of the delivered mechanical power.
We had to reject this solution, as we have not been allowed to hook anything on this part of the turboshaft.
2. On the burning chamber: it is obviously the best place to catch heath.
Unfortunately, we have not been allowed to hook anything on this part of the turboshaft.
3. On the exhaust pipe: this exhaust pipe is a simple mechanical part, quite simple and not very critical. It can thus been easily modified to fit a thermal assembly.
The hot burning gas is flowing in this pipe at quite a high temperature and a fair flow rate. Many analysis and simulations have been held and at the end we had to conclude that scavenging this heath was almost impossible according to the requirements. For several operating points, it can work, but most of the time it can’t. Several reasons for that:
• The gas flow as a typical shape in the pipe and we are not allowed to change anything. So, including anything in the flow is forbidden;
• We can collect heath on the outer skin of the pipe. But the heat transfer between the gas and the metal is very poor, and at the end we cannot meet the requirements;
• The best solution is to lead directly a share of the hot gas flow to the thermo cells, avoiding any thermal interface. We are allowed to use up to 10% of the exhaust gas. It is good enough to deliver the power when the turboshaft is at full power, but does not meet the requirement in other situations.

2.3 Cold source considerations
Several cold sources are available, but none of them were easily usable:
1. Fuel is a very interesting coolant as it always flows at high rate of flow. However, we could easily demonstrate that:
o Temperature rise of the fuel could be unacceptable in specific conditions. However, due to the bleeder regulation used on turbo shafts, this flow can be easily increased to meet an acceptable temperature rise;
o Using the fuel tank as a thermal bin can lead to unacceptable temperature rises (for example when the tank is nearly empty). Specifically because the current fuel circuit and tank has no capability to exchange heat with the outside. Currently, no solution could be found, but we can easily imagine transformations of the fuel circuit to allow such a thermal exchange.
2. In addition to lubrication, oil is used to cool several key elements of the turbo shaft: bearings, combustion chamber skin… It is perfectly designed for such thermal exchange and thus is a perfect candidate as a cold source. Unfortunately, the circuit has no security margin. Very quickly oil temperature is over the maximum acceptable and we cannot use it. Changing the oil grade or any other change has not been accepted and we had to give up this option.
3. Air at the intake of the turboshaft is a very good coolant as there is a high flow rate of fresh air. In addition, this flow is very stable and predictive. Adding an air exchanger is quite easy and very efficient. The problem is that intake is at the front of the turboshaft and the hot source at the aft. Because of the necessary thermal power, we need an unacceptable mass of thermal duct. To use properly this cooling point, we need to take heat close to the front, in the compressor stages. This was not allowed, and we couldn’t use this cold source;
4. Air around the helicopter is also a cold source, but with many drawbacks:
o This fresh air flow is not part of the turboshaft but of the helicopter. It means that a critical function of the turboshaft is coming from the helicopter environment. This is hardly acceptable as this dependence reduces the flexibility of the engine solution;
o A main contributor of this air flow is the rotor shaft. It means that it deeply depends of the pilot actions, and is not really predictive;
o Another main contributor is the speed of the ait around the helicopter: speed of the helicopter itself, but also wind speed around the helicopter. Even if we accept the air speed change, we also have to cope with the direction change. It adds complexity and reduces efficiency of the air exchangers.
o Air behaviour has never been characterized in a correct way to make clear decisions. We had to make assumptions with poor justifications.
At the end, we had no other choice than this solution, have clearly in mind that it was not a safe industrial solution but simply a pragmatic decision to go further in the project.

Potential Impact:
As formerly explained, the project failed to demonstrate this solution on current helicopter engines, mainly because of the engine itself. However, an exhaustive technical analysis has been held and many technical facts are now known and understood. As a consequence:
- the interest of such a system has been proven for this application, and there is clearly a track to follow for future implementations;
- typical performance and constraints have been established, giving a better visibility;
- the key technical points have been featured, and can be used by engine designers to implement such a solution in a future product.

In addition to this typical application, it is clear that this technology has an interesting potential for other similar applications. We have seen that the main technical problem is the temperature range of the thermal sources: if too wide, we have to accept an efficiency penalty. If narrow enough, we can expect a very stable and safe performance. We have identified applications where this thermal spread is narrow and where some energy can have a great impact. The most promising is underwater pipeline for oil and gas industry. These pipelines are located on the sea bed, deep under the surface (down to 5000 meters) and need several remote sensors, safety devices or acoustic beacons. The usual way to feed them is to use an umbilical cable feeding these equipment from the surface power. It is clearly expensive, difficult to maintain and very inefficient as the devices to feed only need a few watts. An interesting way to propose a simple cost effective solution is to use the oil temperature in the pipeline (usually stable around 100°C) and the sea water temperature at 4°C. In that way, we could very efficiently and stably provide enough power.
Several members of the consortium are active on this idea, in conjunction with major oil and gas companies. We are quite confident about a new momentum of the project. A project in a partnership program has been recently proposed.

List of Websites:
For further details, please contact:
Franck Mariette
fmariette@tronico-alcen.com
+33 670 273 386

Final Report Summary - THETAGEN (Thermoelectric generator for engine control system)

Partager cette page

Télécharger