Technology Characterization and Evaluation

NREL researchers developed the Technology Characterization and Evaluation (Tyche) software to help decision makers identify research and development (R&D) investments that can meet goals around technology progress while accounting for uncertainty.

This software will provide a means of evaluating and comparing the impacts of R&D investments on energy technology metrics. Tyche informs R&D decision-making across technologies and projects by providing a general and reproducible workflow that helps analysts identify and evaluate energy technology portfolio risks and benefits and communicate these findings to decision makers.

Tyche provides ensemble simulation and stochastic optimization functionality for evaluating R&D investments according to their potential for contributing toward different program targets and goals. The software encourages collaborators to explore R&D options and develop intuition in decision-making. The quantitative analytical support provided by Tyche can help inform and explain funding decisions by identifying the distribution of R&D investments that optimizes cost and performance improvements under uncertainty. If uncertainties are too large for simulations to provide definitive guidance, Tyche can help identify which input probabilities are the best targets for narrowing uncertainties through additional data collection or expert elicitation.

Capabilities

• Evaluate uncertain improvements in technology metrics that result from R&D investments
• Optimize the distribution of R&D budgets across research projects to meet high-priority targets and goals, while accounting for uncertainty
• Compare metric progress and R&D investments at multiple levels of detail—technology components, subsystems, and systems

Using Tyche

Tyche is written in Python 3.10 and is available as open-source software released on Github under the MIT license. Tyche documentation is also publicly available.

The codebase is comprised of code modules for comparing technology designs, evaluating the impacts of investment scenarios, and performing single- or multi-objective optimization. Tyche users will need to know basic Python coding to build and use custom technology models. Inputs to Tyche consist of Excel spreadsheets containing technology, expert elicitation, and investment data. Example input datasets are provided with the software.

Diagram showing the flow of information through the Tyche software.

Case Studies

Photovoltaic Modules

Specific photovoltaic technology considerations were analyzed using Tyche to model R&D investment estimates to aid in decision-making. In one instance using the software, potential improvements in two parameters within the photovoltaic technology (kerf loss and silicon wafer thickness) were evaluated to determine how investments should be made to improve the levelized cost of energy of the technology. By combining the elicited parameter values and distributions with a bottom-up techno-economic model of photovoltaic module manufacturing, analysts were able to demonstrate the potential range of cost savings achievable under different levels of R&D investment in the two technology parameters.

Biorefinery Technologies

In one case study, five biorefinery technologies using different oily feedstocks to produce the same slate of biofuel co-products were evaluated across different levels of R&D investment. Using Tyche, analysts were able to optimize R&D allocations to achieve the greatest impact on the biofuel minimum selling price. Analysts were also able to determine where they could make the greatest reductions in selling price and in life cycle greenhouse gas emissions of the primary biofuel products.

Related Publications

Methods for R&D Portfolio Analysis and Evaluation, NREL Technical Report (2020)

Energy Sector Portfolio Analysis With Uncertainty, Applied Energy (2022)

Contact