A g-function calculator for Python
Project description
pygfunction: A g-function calculator for Python
What is pygfunction?
pygfunction is a Python module for the calculation of thermal response factors, or g-functions, for fields of geothermal boreholes. g-functions form the basis of many simulation and sizing programs for geothermal heat pump systems. g-Functions are superimposed in time to predict fluid and ground temperatures in these systems.
At its core, pygfunction relies on the analytical finite line source solution to evaluate the thermal interference between boreholes in the same bore field. This allows for the very fast calculation of g-functions, even for very large bore fields with hundreds of boreholes.
Using pygfunction, g-functions can be calculated for any bore field configuration (i.e. arbitrarily positionned in space), including fields of boreholes with individually different lengths and radiuses. For regular fields of boreholes of equal size, setting-up the calculation of the g-function is as simple as a few lines of code. For example, the code for the calculation of the g-function of a 10 x 10 square array of boreholes (100 boreholes total):
import pygfunction as gt
import numpy as np
time = np.array([(i+1)*3600. for i in range(24)]) # Calculate hourly for one day
boreField = gt.boreholes.rectangle_field(N_1=10, N_2=10, B_1=7.5, B_2=7.5, H=150., D=4., r_b=0.075)
gFunc = gt.gfunction.gFunction(boreField, alpha=1.0e-6, time=time)
gFunc.visualize_g_function()
Once the g-function is evaluated, pygfunction provides tools to predict borehole temperature variations (using load aggregation methods) and to evaluate fluid temperatures in the boreholes for several U-tube pipe configurations.
Requirements
pygfunction was developed and tested using Python 3.7. In addition, the following packages are needed to run pygfunction and its examples:
- matplotlib (>= 3.5.1),
- numpy (>= 1.21.5)
- scipy (>= 1.7.3)
- SecondaryCoolantProps (>= 1.1)
The documentation is generated using Sphinx. The following packages are needed to build the documentation:
- sphinx (>= 4.4.0)
- numpydoc (>= 1.2.0)
Quick start
Users - Download pip and install the latest release:
pip install pygfunction
Alternatively, download the latest release and run the installation script:
pip install .
Developers - To get the latest version of the code, you can download the repository from github or clone the project in a local directory using git:
git clone https://github.com/MassimoCimmino/pygfunction.git
Install pygfunction in development mode (this requires pip >= 21.1
):
pip install --editable .
Once pygfunction is copied to a local directory, you can verify that it is
working properly by running the examples in pygfunction/examples/
.
Documentation
pygfunction's documentation is hosted on ReadTheDocs.
License
pygfunction is licensed under the terms of the 3-clause BSD-license. See pygfunction license.
Contributing to pygfunction
You can report bugs and propose enhancements on the issue tracker.
To contribute code to pygfunction, follow the contribution workflow.
Contributors
Massimo Cimmino 💻 📖 💡 :rocket: 🤔 🚧 👀 |
Jack Cook 💻 💡 🤔 📖 |
Matt Mitchell 💻 🤔 |
This project follows the all-contributors specification. Contributions of any kind welcome!
History of changes
Version 2.2.3 (2024-07-01)
New features
- Issue 276 - Added functions to the
boreholes
module for the generation of rectangular fields in a staggered configuration.
Enhancements
- Issue 291 - Simplified the expressions in heat_transfer._finite_line_source_steady_state`. The function is now approximately 25% faster.
Bug fixes
- Issue 255 - Default to an
orientation
of0.
whentilt
is0.
inboreholes.Borehole
class. - Issue 266 - Fixed an issue were
SingleUTube.get_temperature
returned incorrect results when the fluid-to-pipe wall resistance was small in coaxial configurations. New coefficients are introduced inSingleUTube.coefficients_temperature
andSingleUTube.coefficients_outlet_fluid_temperature
. This also solves issues encountered when the fluid mass flow rate is small. - Issue 274 - Fixed scalar assignment from ndim-1 array. It is deprecated as of
numpy
version1.25
. Only ndim-0 arrays can be treated as scalars. - Issue 285 - Use
numpy.complex128
instead ofnumpy.cfloat
. This is to comply with backward-incompatible changes introduced innumpy
version2.0
. - Issue 286 - Fixed incorrect coefficients in
pipes.SingleUTube._continuity_condition_base
which caused errors in all dependent class methods whensegment_ratios
were not symmetric around the borehole mid-length. - Issue 298 - Fixed incorrect coefficients in
pipes._basePipe
,pipes.MultipleUTube
andpipes.IndependentMultipleUTube
which caused errors in fluid temperature profiles and outlet fluid temperatures.
Version 2.2.2 (2023-01-09)
Enhancements
- Issue 204 - Added support for Python 3.9 and 3.10. CoolProp is removed from the dependencies and replace with SecondaryCoolantProps.
Bug fixes
- Issue 231 - Fixed an issue where the evaluation of g-functions at very low times raises an error due a singular matrix. g-Functions below a threshold time value
t=max(r_b)**2/(25*alpha)
are now linearized.
Other changes
- Issue 229, Issue 247 - Added citation to IGSHPA conference paper on pygfunction v2.2 in the documention. Added a
CITATION.cff
file to suggest a correct citation on github. - Issue 230 - Configured github actions to publish pygfunction on Pypi on creation of a release on github.
Version 2.2.1 (2022-08-12)
Bug fixes
- Issue 220 - Fixed the expected line length in
boreholes.field_from_file()
to correctly import fields of inclined boreholes. - Issue 224 - Fixed an issue where tests were not run on maintenance branches.
Version 2.2.0 (2022-07-10)
New features
- Issue 50 - Implemented inclined boreholes for the evaluation of g-functions. The implementation includes an approximation of the FLS solution for inclined boreholes based on the method of Cimmino (2021) (see Issue 138). The
'equivalent'
solver is not yet supported. - Issue 138 - Implemented the approximation of the finite line source solution of Cimmino (2021). The approximation avoids the numerical evaluation of integrals. This speeds up the calculation of g-functions when enabled.
- Issue 148 - Implemented
effective_borehole_thermal_resistance()
andlocal_borehole_thermal_resistance()
methods for all pipe classes. Deprecatedpipes.borehole_thermal_resistance()
, which computed the effective borehole thermal resistance. It will be removed inv3.0.0
. Implemented a newupdate_thermal_resistances()
method to all pipe classes. This method allows to update the delta-circuit of thermal resistance of the boreholes based on provided values for the fluid thermal resistances. This allows simulations with time-variable fluid thermal resistances.
Enhancements
- Issue 152 - Vectorized
coefficients_temperature
and_general_solution
inpipe
objects to accept depthsz
as an array. This speeds up calculations forget_temperature
andget_borehole_heat_extraction_rate
class methods. - Issue 183 - Vectorized
pipes.multipole()
andpipes._Fmk()
to decrease the calculation time ofpipes.thermal_resistances()
. Amemoization
technique is implemented to reduce computation time for repeated function calls to further speed-up the initialization ofPipe
andNetwork
objects. - Issue 198 - Refactored the
'detailed'
solver to evaluate same-borehole thermal response factors in a single call tofinite_line_source_vectorized()
. This speeds up calculations of g-functions using the'detailed'
solver. - Issue 199 - Changed the integral bounds to avoid repeated evaluation of integrals over semi-infinite intervals. This speeds up calculations of g-functions using all solvers and the evaluation of the finite line source solution with
time
as an array. - Issue 206 - Refactored
boreholes.find_duplicates()
to usescipy.spatial.distance.pdist()
for the calculation of distances between boreholes. This leads to faster initialization of thegFunction
class for large borefields.
Other changes
- Issue 80 - Added references to the
pipes
module for the evaluation of borehole thermal resistances. - Issue 171 - Refactored modules and examples to use the built-in
enumerate(x)
instead ofrange(len(x))
. - Issue 172 - Refactored reports of calculation time to use
time.perf_counter()
instead oftime.time()
. - Issue 173 - Refactored strings into f-strings instead of using
str.format()
. - Issue 177 - Converted
setup.py
script tosetup.cfg
andpyproject.toml
files. This is motivated by PEP518 and PEP621. - Issue 179 - Refactored tests to use the
pytest
package instead ofunittests
. - Issue 180 - Configured
tox
and github actions for continuous integration.
Bug fixes
- Issue 192 - Fixed comparison of
time
withnumpy.inf
inheat_transfer.finite_line_source
that caused the function to fail whentime
is an array. - Issue 193 - Fixed
heat_transfer._finite_line_source_integrand
,heat_transfer._finite_line_source_equivalent_boreholes_integrand
, andheat_transfer._finite_line_source_steady_state
to return an array of zeros of the expected shape whenreaSource==False and imgSource==False
. - Issue 196 - Fixed "invalid escape sequence" warnings when running tests on github actions.
- Issue 202 - Added missing package
recommonmark
to requirements for documentation and development. - Issue 208 - Fixed an issue where
boreholes.field_from_file()
failed when the text file only contained 1 borehole.
Version 2.1.0 (2021-11-12)
New features
- Issue 36 - Added a
Coaxial
class to thepipes
module to model boreholes with coaxial pipes. - Issue 135 - Added functionality for non-uniform discretization of the segments along the boreholes. This increases the accuracy of g-function calculations for the same number of segments when compared to a uniform discretization. Segment lengths are defined using the
segment_ratios
option in thegFunction
class. Adiscretize
function is added to theutilities
module to generate borehole discretizations using an expanding mesh. - Issue 146 - Added new solver
'equivalent'
to thegFunction
class. This solver uses hierarchical agglomerative clustering to identify groups of boreholes that are expected to have similar borehole wall temperatures and heat extraction rates. Each group of boreholes is represented by a single equivalent borehole. The FLS solution is adapted to evaluate thermal interactions between groups of boreholes. This greatly reduces the number of evaluations of the FLS solution and the size of the system of equations to evaluate the g-function.
Enhancements
- Issue 118 - Refactored checks for stored
_BasePipe
andNetwork
coefficicients to usenumpy.all()
. This decreases calculation time. - Issue 119 - Refactored
Network
class to change how coefficient matrices are calculated. This decreases calculation time. - Issue 132 - Refactored
SingleUtube
andMultipleUTube
classes to eliminatefor
loops in the calculation of matrix coefficients. This decreases calculation time whennSegments>>1
. - Issue 133 - The
nSegments
argument is now able to take in the number of segments for each borehole as a list. Each borehole must be split into at least 1 segment, and the length of the segment list must be equal to the number of boreholes. - Issue 141 - Changed the calculation of the convective heat transfer coefficient in the transition region (
2300. < Re < 4000.
) byconvective_heat_transfer_coefficient_circular_pipe()
. The Nusselt number is now interpolated between the laminar value (atRe = 2300.
) and the turbulent value (atRe = 4000.
). This avoids any discontinuities in the values of the convective heat transfer coefficient nearRe = 2300.
.
Other changes
- Issue 93 - Reformatted
pipes
andnetworks
modules to use the@
matrix product operator introduced in PEP465. This improves readability of the code. - Issue 100 - Replaced calls to
numpy.asscalar()
with calls toarray.item()
.numpy.asscalar()
is deprecated as ofnumpy
version1.16
. - Issue 124 - Reformatted
pipes
andnetworks
modules to clarify the expected values form_flow
parameters. These are replaced by any ofm_flow_pipe
,m_flow_borehole
orm_flow_network
depending on the function or class method. Added a nomenclature of commonly used variables to the documentation. - Issue 125 - Refactored class methods and docstrings in
Pipe
andNetwork
objects to better represent the expected shapes of array inputs and outputs. - Issue 139 - Updated requirements for numpy from version
1.19.2
to1.20.1
. Clarified the Python version3.7
requirement in theREADME.md
file. - Issue 154 - Replaced
numpy.int
andnumpy.bool
dtypes in array initializations with built-in typesint
andbool
.numpy.int
andnumpy.bool
are deprecated as ofnumpy
version1.20
. - Issue 158 - Changed default parameter values for g-function calculations. The
gFunction
class now uses the'equivalent'
solver by default and a non-uniform discretization ofnSegments=8
given byutilities.segment_ratios()
. - Issue 160 - Deprecated functions
gfunction.uniform_heat_extraction
,gfunction.uniform_temperature
,gfunction.equal_inlet_temperature
andgfunction.mixed_inlet_temperature
. They will be removed inv3.0.0
.
Bug fixes
- Issue 99 - Fixed an issue where
MultipleUTube._continuity_condition()
andMultipleUTube._general_solution()
returned complex valued coefficient matrices. - Issue 130 - Fix incorrect initialization of variables
_mix_out
and_mixing_m_flow
inNetwork
. - Issue 155 - Fix incorrect initialization of variables in
Network
and_BasePipe
. Stored variables are now initialized asnumpy.nan
instead ofnumpy.empty
. - Issue 159 - Fix
segment_ratios
function in theutilities
module to always expect 0 <end_length_ratio
< 0.5, and allows fornSegments=1
ornSegments=2
. If 1<=nSegments
<3 then the user is warned that theend_length_ratio
parameter is being over-ridden.
Version 2.0.0 (2021-05-22)
New features
- Issue 33, Issue 54, Issue 85 - New class
gFunction
for the calculation of g-functions. The new class is a common interface to all boundary conditions and calculation methods. The new implementation of the solver reduces the memory requirements of pygfunction. The new class implements visualization features for the g-function and for heat extraction rates and borehole wall temperatures (both as a function of time and for the profiles along the borehole lengths). - Issue 75 - New module
media
with properties of brine mixtures. - Issue 81 - Added functions to find a remove duplicate boreholes.
Enhancements
- Issue 78, Issue 109 - Optimization of solvers for the calculation of g-functions. The finite line source (FLS) solution is now calculated using
scipy.integrate.quad_vec
which significantly improves calculation time overscipy.integrate.quad
. The identification of similarities in the 'similarities' solver has also been refactored to identify similarities between boreholes as an intermediate step before identifying similarities between segments. The calculation time for the identification of similarities is significantly decreased. - Issue 94 - Refactor visualization functions and methods to uniformize figure styles across modules.
- Issue 108 - Optimize the load aggregation algorithm of Claesson and Javed using
numpy.einsum()
. - Issue 112 - Optimization of
_BaseSolver.temporal_superposition()
. The computationally expensive for loop is replaced by a call tonumpy.einsum()
. This decreases the calculation time of large bore fields. - Issue 114 - Optimization of
_finite_line_source_integrand()
. The call to_erfint()
is now vectorized. This decreases the number of calls by a factor 8 during integration. The calculation time of g-functions is decreased, especially for smaller bore fields.
Bug fixes
- Issue 86 - Documentation is now built using Python 3 to support Python 3 features in the code.
- Issue 103 - Fixed
gFunction
class to allow both builtinfloat
andnumpy.floating
inputs. - Issue 104 - Raise an error if g-function is calculated with inclined boreholes. This will be supported in a later version of pygfunction.
Other changes
- Issue 72 - Added a list of contributors to the front page. The list is managed using all-contributors.
- Issue 87 - Drop support for Python 2. All package requirements are updated to the latest conda version.
- Issue 96 - Added a reference to the conference paper introducing pygfunction in the documentation.
Version 1.1.2 (2021-01-21)
New features
- Issue 47 - Added verification of the validity of pipe geometry to pipe classes. Extended visualization of the borehole cross-section.
- Issue 66 - Added a class method to the Claesson & Javed load aggregation method to retrieve the thermal response factor increment.
Enhancements
- Issue 59 - Use a relative tolerance instead of an absolute tolerance in the identification of borehole pair similarities. This provides faster execution times and similar accuracy.
Bug fixes
- Issue 58 - Store matrix coefficients in
Network
class methods for re-use when inlet conditions are constant. - Issue 64 - Fixed an issue where the g-function was returned as an array of integers if time values were integers.
Version 1.1.1 (2020-06-20)
New features
- Issue 40 - Added Network class for simulations involving networks of boreholes.
Bug fixes
- Commit a4f6591 - Fixed import of Axes3D necessary in
borehole.visualize_field()
.
Version 1.1.0 (2018-03-09)
New features
- Commit 2bd12bd - Implemented
UTube.visualize_pipes()
class method. - Issue 30 - Laminar regime is now considered for calculation of convection heat transfer coefficient.
- Issue 32 - g-Functions for bore fields with mixed series and parallel connections between boreholes.
Bug fixes
- Commit 2523f67 -
boreholes.visualize_field()
now returns the figure object. - Issue 25 - Fixed documentation of ./examples/uniform_temperature.py.
- Issue 27 -
thermal_response_factors()
is now part of theheat_transfer
module (moved fromgfunction
). - Commit 1f59872 - Fixed incorrect summation limit in pipes._F_mk.
Version 1.0.0 (2017-12-01)
New features
- Issue 4 - Unit testing and integration with Travis CI.
- Issue 16 - Added capability to import bore field from external text files.
- Issue 18 - Added capability to visualize bore fields.
- Issue 20 - Added utilities module.
- Issue 5 - Added setup.py installation script using setuptools. pygfunction will now be available on pypi.
Bug fixes
- Commit 0b2d364 -
boreholes.U_shaped_field()
andboreholes.U_shaped_field()
did not construct the field properly when called with low form factors - Commit 62acd42 - Criteria for smooth pipe correlation in calculation of Darcy friction factor was too large. Colebrooke-White equation is now always used for turbulent flow.
- Issue 14 - Evaluate g-functions at a single time value.
- Commit 7b408e9 - Fix interpolation of thermal response factors in cases where the minimum timestep does not correspond to the first timestep. This also fixes errors caused by rounding errors in the interpolation.
- Issue 22 - Allow 1d arrays for g-function values in the initilization of load aggregation algorithms.
Version 0.3.0 (2017-10-17)
New features
- Issue 6 - Store coefficients in pipe models for faster computation
- Issue 7 - Multipole method to evaluate borehole internal resistances
Copyright (c) 2017-2024, Massimo Cimmino All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
-
Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
-
Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
-
Neither the name of pygfunction nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file pygfunction-2.2.3.tar.gz
.
File metadata
- Download URL: pygfunction-2.2.3.tar.gz
- Upload date:
- Size: 108.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.1.0 CPython/3.12.4
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | b0e73bf66c7e3921b957459edf859a96c93ad8e0f6f30c07bb251f82d9f8264e |
|
MD5 | 50014eeccaa62daf3a40db247c52e343 |
|
BLAKE2b-256 | cf95bde101951aeeba5d7e61c421348c91eef2516e5470d675481ac1f7ab4548 |
File details
Details for the file pygfunction-2.2.3-py3-none-any.whl
.
File metadata
- Download URL: pygfunction-2.2.3-py3-none-any.whl
- Upload date:
- Size: 97.3 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.1.0 CPython/3.12.4
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 0b1219a55b7b6a4ce9f5197f4df20f16b10b29276634b2e1c1dea264192ed1fc |
|
MD5 | ae244faf8175431f2ea03df7672a00bf |
|
BLAKE2b-256 | 676fd25d575b4f255deb363ec8b12768ce7615e70fe185af6bb7f7d83cec100c |