CS15: Digitalization and Energy Markets

The transition toward low-carbon energy systems requires mechanisms that enable the large-scale deployment of renewable fuels while preserving environmental integrity and compatibility with existing infrastructure. In liquid fuel supply chains, renewable and fossil fuels are frequently blended and distributed through shared logistics networks, making physical traceability of renewable content impractical. Book-and-claim mechanisms have emerged as a viable accounting approach to decouple physical fuel flows from sustainability attributes, allowing environmental benefits to be allocated independently of fuel delivery. This work develops a blockchain-based book-and-claim framework for Diesel R, a diesel fuel with renewable content produced by Petrobras through the coprocessing of mineral diesel and renewable raw materials. The proposed system leverages blockchain to enable the secure issuance, transfer, and retirement of digital certificates representing the renewable attributes of Diesel R. The methodology starts with tokenization. Based on the verified sustainable volume, Environmental Attribute Tokens (EATs) are minted on the Cardano blockchain. Each EAT represents one cubic metre (cbm) of certified sustainable content. These tokens are transferred to the diesel producer, who controls the owner wallet and holds the primary rights over the environmental attributes generated during production. Subsequently, a purchasing company acquires EATs from the producer or through a secondary market. The transfer of EATs occurs independently of the physical delivery of diesel, in line with the B&C principle that decouples product flows from sustainability attributes. After acquiring the tokens, the purchasing company performs the claim, also referred to as the inset, by retiring the EATs to substantiate its sustainability or emissions reduction claims. The study contributes by extending book-and-claim applications to renewable diesel fuels in the Brazilian context and by demonstrating how blockchain-based systems can support scalable, transparent, and reliable attribution of renewable fuel attributes without relying on physical segregation.

Electricity markets in many industrialized nations are characterized by high dynamism, driven by energy and climate policy decisions, geopolitical changes, and resource scarcity. Simultaneously, increasing digitalization, also shaped by new AI applications, is increasing competitive pressure on energy service providers to generate and strengthen brand loyalty.
An important question is, what brands or services are selected by customers in any given choice situation and if there are influencing factors for the purchase decision process on electricity markets. Our approach is one of the few that relies entirely on consumption goods in full-access markets, where understanding, influencing, and modeling customer buying behavior is a concern for businesses and sales managers. We are investigating the influence of a context effect on the reliability of preferences. This effect assumes that creating interaction with a topic before a purchase decision (group one) results in a different preference estimate than in a purchase decision in which no such interest was created (group two). We test if preferences are subject to change due to situational context prior to the buying decision, dependent on digital maturity. 
The use of attitude questions prior to the actual choice task in this application leads to a different willingness to pay (WTP) for the price calculation elements, particularly in favor of the more digitized and technologically advanced variants. Both groups experience disutility from the deviation of a fixed price per kWh. Also, service bundles are less attractive if they are based on consumption-based discounts or premiums in case of higher vs. lower electricity consumption.
These findings show that a treatment that addresses digital attitudes does not necessarily induce a higher WTP for digital attributes apart from pricing-related items, leading to the result that a treatment is of little use to promote novel electricity contracts, except the novelty is price-related.

Modern economies are experiencing an unprecedented expansion in data generation and analytics. This data revolution enables more efficient processes and optimized infrastructure management. Yet, concerns about the boundaries of privacy and the value individuals assign to their personal information have intensified. This creates a central policy challenge: while data improve system performance, privacy risks may discourage individuals from sharing information or supporting digital energy technologies.
Although prior research documents widespread privacy concerns, less is known about whether privacy-seeking behavior is causally driven by individuals' valuation of their personal data. If households underestimate the inferential power of smart meter data, their privacy valuation (and related behavior) may be formed under misperceptions. While information provision can update beliefs, it remains unclear whether belief-induced changes in valuation translate into concrete privacy-seeking intentions.
To address this gap, we piloted a randomized information treatment with an instrumental variable strategy to causally identify the effect of post-treatment data valuation on intended privacy-seeking behavior. First, respondents report their beliefs about how accurately smart meter data reveal daily routines and state their valuation of hourly consumption data. Then, the treatment group receives factual information that predictive accuracy exceeds 90% before reassessing their beliefs and valuation. Finally, to identify the causal effect of valuation on behavior, we instrument post-treatment valuation with the interaction between pre-treatment belief error and the treatment indicator in a two-stage least squares framework.
Instrumental variable estimates show that higher privacy valuation causally increases intended privacy-seeking behavior. A one-euro increase in post-treatment valuation raises the probability of intending to provide less granular data by 0.8 percentage points and checking privacy settings by 0.7 points. These findings demonstrate that privacy-seeking behavior is valuation-driven and responsive to corrected misperceptions, with implications for data governance and digital energy policy.

Bunker fuel oils are usually referred to as heavy residual fuels in ships, which represents a high impact on greenhouse gas (GHG) emissions and, consequently, climate change. Because of that, regulatory bodies have been increasingly establishing standards for cleaner fuels, which may play an important role in decarbonization and energy transition through sustainable shipping fuels such as LNG, hydrogen, ammonia, and biofuels. Those sustainable fuels may be blended with fossil fuels on the final ship fuel but may also be used at different stages at their production and transport supply chains. In any of those scenarios, it may help in efforts of decarbonization, and because of that it must be tracked and accounted. This work develops a blockchain-based Chain of Custody framework for bunker fuels with renewable content produced by Petrobras. Chain of Custody (CoC) is a chronological physical or electronic proof of purchase, transport, control, transfer and use of a product or certain product properties. It certifies that specified standards are adhered to throughout the entire supply chain. In this context, the proposed framework uses blockchain technology to enable the secure issuance, transfer, and retirement of digital certificates representing the renewable attributes of sustainable bunker. This is digitally implemented on the Cardano blockchain, where tokens representing the sustainable fuel components are transferred along the supply chain through the blockchain wallet of the current fuel owner. This work contributes by extending Chain of Custody applications to renewable bunker fuels and by demonstrating how blockchain-based systems can support scalable, transparent, and reliable attribution of renewable fuel attributes without relying on physical segregation.