(436d) Mapping Environmental and Economic Analysis of Decentralized Cogeneration Energy Management Centers | AIChE

(436d) Mapping Environmental and Economic Analysis of Decentralized Cogeneration Energy Management Centers


In the midst of a global conversation on how to reduce GHG emissions, Ontario's electrical grid stands out due to its extensive use of low carbon-intense processes such as nuclear and hydropower that acts as baseload energy sources. However, for peak electricity demands that require a quick ramp up of production, natural gas (NG) turbines are deployed [1], which causes the carbon intensity of Ontario’s electrical grid to vary hourly [2]. Another intensive and permanent Canadian energy requirement is heating. Currently, it is provided mostly by NG -fired boilers [3]. The pursuit for better NG utilization is the motivation for this study, which proposes a design of combined heat and power energy management center (CHP-EMC).

CHP-EMC is composed of an NG-powered internal combustion engine connected to an electrical generator (NG-CHP), a hot water tank storage, an NG boiler, and a cooling tower to keep the water temperature within acceptable bounds. EMC is connected to a water loop similar to the ones used in district heating (DH) by a heat exchanger and responds to a community's heating demand using an hourly lumped-demand approach. The simulations are developed in TRNSYS, which is a capable heat transfer and fluid dynamic simulator [4]. The storage tank uses a built-in dynamic model whilst the rest of the components operate in pseudo steady-state. These components are connected by pipe models that account for time delay and flowrate.

The equipment presence and its sizes are decided by minimizing the natural gas consumption while simultaneously maximizing thermal and electrical generation. The optimization strategy used is a black-box approach using modeFrontier software [5]. This optimization and data visualization software can be coupled with any other software that uses input/output files -such as TRNSYS- and it executes several simulations guided by the optimization strategy and its constraints. For this work, classic Multi-Objective Genetic Algorithm (MOGA) is chosen for design variables (including the presence of specific components) and the pareto points are analyzed in more detail. TRNSYS simulation has internal simple controllers that take care of schedule and operational constraints. The result of the optimization aims to create an EMC flexible enough to endure deviations from expected demand profiles.

The emissions of proposed EMC designs are compared to the status quo (electricity supplied by electrical grid and thermal demand supplied by NG boilers) for a year of operation. Total emissions are calculated for each possible value of gCO2eq/kWh to better represent possible scenarios of carbon intensity. The results are summarized in an easy-to-read emissions vs technology map that defines the conditions under which the EMC is less pollutant than status quo (SQ). The optimal design and selection of the EMC (vs. SQ) system is strongly dependent on the carbon intensity of the electric grid during peak-load times, the electric demand profiles, and the cooling demand profiles over the course of its operation. These are best characterized by the combination of greenhouse gas emissions per kWh during peaking operation, the ranges of heating and electricity demand over the course of the year (the maximum minus the minimum), and the kurtosis of the demands.

Following this analysis, an economic evaluation of both systems is carried out, demonstrating the cost of carbon avoided (CCA). CCA is a metric that compares CO2 and cost of two technological strategies and calculates the cost per ton of GHG emissions potentially reduced [6]. These results are compared to possible carbon tax values, in a similar way to carbon intensity. A second map, cost vs technology, is then developed for easy visualization of the impact of carbon taxes on EMC and SQ operation costs.

The goal of this work is to apply the benefits of decentralized EMCs to reduce GHG emissions from both heating and electricity generation. The results are generalizable and can be applied to any location due to its methodology in representing carbon intensity and carbon tax.


[1] IESO, "Power Data," 30 04 2021. [Online]. Available: https://www.ieso.ca/power-data. [Accessed 30 04 2021].

[2] L. Pereira and I. D. Posen, "Lifecycle greenhouse gas emissions from electricity in the province of Ontario at different temporal resolutions," Journal of Cleaner Production, vol. 270, no. 122514, 2020.

[3] S. Canada, "Primary heating systems and type of energy," 03 04 2021. [Online]. Available: https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=3810028601. [Accessed 28 04 2021].

[4] TRNSYS, 2019. [Online]. Available: http://www.trnsys.com/index.html. [Accessed 30 04 2021].

[5] ESTECO, "modeFrontier," 2021. [Online]. Available: https://engineering.esteco.com/modefrontier/. [Accessed 30 04 2021].

[6] K. Gillingham and J. H. Stock, "The Cost of Reducing Greenhouse Gas Emissions," Journal of Economic Perspectives, 2018.