Performance Curves Series: Part 2 – Key Fundamentals & Foundations (Working Draft)

Performance Curves Series: Part 2 – Key Fundamentals & Foundations (Working Draft)

Part 2 of this series is structured as a series of prompts/questions and answers. Depending on your personal experience, some or all of this may seem rudimentary, but I want to ensure a few things are covered comprehensively before building on a shaky foundation. I welcome any suggestions for additions to this section, if you personally feel anything is not covered that needs to be here!

  • What is Coefficient Regression?
    • Curve Fits – explain illustrate through Excel charts
    • In school, much of the mathematics I was taught (and subsequently forgot) was centered around observing a curve, then determining a formula that would represent that curve. If you can look at a chart and recognize a visual pattern that should fit a formula, then there are methods to working out (solving for) the coefficients in that formula.
    • Excel trivializes this exercise through trendlines, at least for some simpler equation types. For example, if we have the following measured data for XXX, [insert illustrated example], we can use a scatterplot to play with different polynomial equations to determine a “best fit.”
    • To determine (regress) coefficients for more complex and multivariate equations, we need to expand our toolbox. Excel happens to contain everything we need to perform and evaluate more complex coefficient regressions. I will be teaching this by way of illustrated example later, in Part 5 of this series.
  • What are Performance Curves?
    • In the context of a building energy simulation model: Performance Curves are components of a model which allow for complex interactions to estimate the performance/output of a system or equipment, based on one or more variables.
    • Hereafter, I will be specifically referencing these variables as:
      • Dependent Variables – The output of a performance curve
      • Independent Variable – The input(s) of a performance curve, Information needed to produce the Dependent Variable
    • For relatively complex cases, Independent variables (curve inputs) may actually be the Dependent variable (curve output) from other performance curves: (insert visual – chillers in doe2):
  • Some necessarily platform-specific prompts should be addressed. I can provide answers specific to the eQuest/doe2 platform, but if you are using another energy simulation platform, you would be well advised to consider and ensure you can answer the following prompts for that platform. (I would be happy to include/append responses assembled by others):
    • Understand Performance Curve Naming Conventions
    • Identify Viable Performance Curve Equation Types for a given application
    • How can I find, review, manipulate, and create Performance Curves within my platform? What do curves and their inputs look like in my platform?
    • How can I determine viable Performance Curve types for a given piece of equipment/process in my model?


Top