Modelling an electrolyser in a graph-based framework

Buu-Van Nguyen, Johan Romate, Cornelis Vuik · 2024

Read Modelling an electrolyser in a graph-based framework by Buu-Van Nguyen, Johan Romate, Cornelis Vuik (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-02716.txt.

Introduction

Buu-Van Nguyen, Johan Romate, and Cornelis Vuik model an electrolyser as a node in a wider energy network, in a 2024 paper. Electrolysers split water into hydrogen and heat using electricity, linking power, gas, and heating systems. Network planners need a steady-state model of the device that plays well with load-flow calculations for the whole web. The authors write the device into a graph-based framework, where each plant or pipe is an edge or node.

They work out which boundary conditions make the model's equations have exactly one answer. Two cases are covered: the split between hydrogen and heat output fixed in advance, and the split left free. Computer checks confirm the conditions behave as derived. A sound electrolyser block is a small but needed brick for planning the multi-energy systems of the hydrogen era.

Catalog record

CREATORBuu-Van Nguyen, Johan Romate, Cornelis Vuik
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE34 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-02716

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Modelling an electrolyser in a graph-based framework Buu-Van Nguyen, Johan Romate, Cornelis Vuik Abstract We propose an electrolyser model for steady-state load flow analysis of multi-carrier energy networks, where the electrolyser is capable of producing hydrogen gas and heat. We show that there are boundary conditions that lead to a well-posed problem. We derive these conditions for two cases, namely with a fixed and non-fixed ratio between gas and heat output. Furthermore, the derived conditions are validated numerically. [orcid=0000-0002-2810-064X] [orcid=0000-0002-5988-9153] keywords Electrolyser ,Integrated energy network ,Multi-carrier energy systems ,Steady-state load flow analysis ,Well-posedness † † credit: Conceptualization, Methodology, Software, Writing - Original Draft † † credit: Writing - review & editing, Supervision † † credit: Writing - review & editing, Supervision † † corresponding: Corresponding author 1 Introduction The current level of greenhouse gas emissions leads to a significant contribution towards global warming [ 10 ] [ 9 ] . A straightforward solution to global warming is to reduce the level of emissions. This can be achieved by replacing fossil fuels partly with renewable energy sources, such as solar and wind energy [ 1 ] [ 6 ] . The downside of these sources is that it can lead to an unstable power system [ 8 ] [ 2 ] [ 18 ] [ 15 ] . The instability is caused by the variable energy production, due to the dependence on the weather. Fortunately, it is possible to prevent instability of the energy system by using electrolysers. An electrolyser can convert a surplus of electricity into hydrogen gas and heat. While an insufficient amount of energy is produced, the hydrogen gas produced by electrolysers can be utilised to offset the deficiency. However, utilising an electrolyser efficiently, requires a careful analysis on the placement and quantity of electrolysers in the energy network. The analysis is usually done by modelling the energy transport, also known as a load flow analysis. The difficulty in modelling the effect of an electrolyser accurately, is that you have to model multiple energy carriers. In our case, those are electricity, gas and heat, which results in an integrated energy network. A way to model multi-carrier energy networks is by using a graph-based framework [ 13 ] [ 11 ] [ 12 ] . The graph-based framework allows modelling a variety of coupling units, but an electrolyser has not been well researched in this framework. The closest coupling unit is a Power-to-Gas (P2G) unit and an electrical boiler, but it only captures conversion to 1 energy carrier, whilst an electrolyser outputs to different energy carriers. A combined heat and power plant (CHP) is another coupling unit that outputs to different energy carriers. Although, the input energy is different from an electrolyser, which is gas. In this paper, we show a linear model for an electrolyser that can be utilised for steady-state load flow analysis. We show that boundary conditions exist that lead to a well-posed problem. Moreover, we include some numerical experiments to illustrate our approach. 2 Model We model an electrolyser that can convert electricity into gas and residual heat, where we are interested in two cases. The first case considers an electrolyser where it converts gas and heat with a known output ratio. The second case considers a situation where the output ratio has to be determined depending on the energy transport surrounding the electrolyser. Before we start with the model equations, we need a graph representation of an electrolyser. Therefore, we describe how an energy network can be transformed into a graph. A graph consists of nodes connected by links. Each node and link corresponds to an energy network element. For example, a node can represent a source and a link can represent a transmission line. Moreover, each node and link corresponds to a physical law. Explicitly, the nodes are associated with conservation laws and the links are associated with the physical model of the underlying network element. Table 1 shows an overview of the conservation laws and common models for each single-carrier. The models for each single-carrier represents the following: the electrical network is an AC system [ 17 ] [ 11 ] , the gas network is a low-pressure system [ 14 ] [ 11 ] and the heat network is a closed-loop system [ 11 ] . Table 1: Models for electricity, gas and heat networks. Network element Description Model Electricity Node Kirchhoff’s law for active power P_{i}=-\sum_{j}P_{ij} Node Kirchhoff’s law for reactive power Q_{i}=-\sum_{j}Q_{ij} Link Short transmission line (send, P) P_{ij}=g_{ij}|V_{i}|^{2}-|V_{i}||V_{j}|\left(g_{ij}\cos{\delta_{ij}}+b_{ij}\sin{\delta_{ij}}\right) Link Transmission line (send, Q) Q_{ij}=-b_{ij}|V_{i}|^{2}-|V_{i}||V_{j}|\left(g_{ij}\sin{\delta_{ij}}-b_{ij}\cos{\delta_{ij}}\right) Link Transmission line (receive, P) P_{ji}=g_{ij}|V_{j}|^{2}-|V_{i}||V_{j}|\left(g_{ij}\cos{\delta_{ij}}-b_{ij}\sin{\delta_{ij}}\right) Link Transmission line (receive, Q) P_{ji}=-b_{ij}|V_{j}|^{2}+|V_{i}||V_{j}|\left(g_{ij}\sin{\delta_{ij}}+b_{ij}\cos{\delta_{ij}}\right) Gas Node Conservation of mass q_{i}=\sum_{j}q_{ij} Link Pipe p_{i}-p_{j}=(C^{g})^{-2}f|q_{ij}|q_{ij} Heat Node Conservation of mass m_{i}=\sum_{j}m_{ij} Node Conservation of energy (supply) \sum_{l}m_{i,l}T^{s}_{i,l}=\sum_{j}m_{ij}T^{s}_{ij} Node Conservation of energy (return) \sum_{l}m_{i,l}T^{r}_{i,l}=\sum_{j}m_{ij}T^{r}_{ij} Link Pipe p_{i}-p_{j}=(C^{h})^{-2}f|m_{ij}|m_{ij} Link Pipe heat loss (supply) T^{s}_{ji}=T^{a}+e^{=\frac{h\pi DL}{C_{p}m}}\left(T^{s}_{ij}-T^{a}\right) Link Pipe heat loss (return) T^{r}_{ij}=T^{a}+e^{=\frac{h\pi DL}{C_{p}m}}\left(T^{r}_{ji}-T^{a}\right) Terminal link Total heat power \Delta\varphi_{i,l}=C_{p}m_{i,l}(T^{s}_{i,l}-T^{r}_{i,l}) Since an electrolyser interacts with different energy-carriers, it is modelled as a node connected with links of the relevant energy carriers. We assume that these links have no energy losses. The coupling node, that represents an electrolyser, has a model that governs the energy balance given by Equation ( 1 ): \displaystyle\eta P=\text{HHV}q+\Delta\varphi (1) where \eta\in[0,1] is the efficiency, P is the active power, q is the gas flow, HHV is the higher heating value of gas and \Delta\varphi is the heat power. This model is based on the CHP model from [ 16 ] [ 19 ] , where we have adjusted our model based on the input and output energy. However, a CHP can generate electricity and heat in a flexible manner, whilst for an electrolyser only residual heat is available. This behaviour is reflected by Equation ( 2 ): \displaystyle\eta_{h}\eta P=\Delta\varphi (2) where \eta_{h}\in[0,1] is the heat efficiency. In other words, Equation ( 2 ) tells us that a fraction of the available energy is converted to heat. Henceforth, we require two equations to model an electrolyser that converts electricity into gas and heat. Our model allows the electrolyser to produce gas and heat in a flexible manner by letting \eta_{h} to be unknown. Hence, our model is capable of modelling both cases we mentioned in the beginning of this section. From a mathematical point of view, we seek to know when the model is well-posed for these cases. Therefore, we show necessary conditions for when this holds in Section 3 . 3 Boundary conditions In general, integrated energy networks lead to a system of nonlinear equations. These systems are usually solved numerically. Solving a nonlinear energy system is conventionally done with the Newton-Raphson (NR) method. In this method a linearisation of the system of equations is involved, where one has to solve a linear system in order to solve the original system. With regards to linear systems, a necessary condition for well-posedness is that the system of equations is square. Usually, energy network models have more variables than equations, unless we allow additional conditions in the form of specifying variables. In this paper, we denote them as boundary conditions. A side note, nodes with different types of boundary conditions are known as node types in literature related to

Read the complete text