diff --git a/src/tutorial/runtime/configuration-consolidation/index.rst b/src/tutorial/runtime/configuration-consolidation/index.rst
index 8b26c8a33b..79e4e26408 100644
--- a/src/tutorial/runtime/configuration-consolidation/index.rst
+++ b/src/tutorial/runtime/configuration-consolidation/index.rst
@@ -184,4 +184,4 @@ Which Approach To Use
 
 .. ifslides::
 
-   Next section: :ref:`Rose Tutorial <tutorial-rose-configurations>`
+   Next section: :ref:`Reinstalling workflows <tutorial.Reinstalling-workflows>`
diff --git a/src/tutorial/runtime/index.rst b/src/tutorial/runtime/index.rst
index ed68447ac9..4cd09f9e64 100644
--- a/src/tutorial/runtime/index.rst
+++ b/src/tutorial/runtime/index.rst
@@ -17,3 +17,4 @@ This section covers:
    introduction
    runtime-configuration
    configuration-consolidation/index
+   reinstalling
diff --git a/src/tutorial/runtime/reinstalling.rst b/src/tutorial/runtime/reinstalling.rst
new file mode 100644
index 0000000000..c4b7281a75
--- /dev/null
+++ b/src/tutorial/runtime/reinstalling.rst
@@ -0,0 +1,160 @@
+.. _tutorial.Reinstalling-workflows:
+
+Fixing and Reloading Workflows
+==============================
+
+.. admonition:: Aim
+
+
+... note::
+
+   This tutorial explores one working practice. You might, alternatively, wish
+   to use ``cylc shutdown``, ``cylc reinstall`` and ``cylc play`` to modify and restart
+   an existing workflow.
+
+
+
+.. ifnotslides::
+
+   Consider a workflow where a a large part of the runtime has completed,
+   but a late task has failed.
+
+.. digraph:: workflow
+
+   rankdir="LR"
+   get_data -> process_data
+   get_data[
+      label="get data\n(very slow)",
+      color="forestgreen",
+      fillcolor=palegreen,
+      style=filled
+   ]
+   process_data[
+      label="process data",
+      color="red",
+      fillcolor=palevioletred,
+      style=filled
+   ]
+
+.. ifnotslides::
+
+   Having identified the problem with the late task we want to
+   update the runtime definition of the installed workflow rather than
+   re-run the whole thing.
+
+
+.. practical::
+
+   .. rubric:: In this practical we will break the
+      :ref:`weather-forecasting workflow <tutorial-datetime-cycling-practical>`
+      from the :ref:`scheduling tutorial <tutorial-scheduling>`.
+
+   #. **Get your own copy of the workflow**
+
+      Use ``cylc get-resources`` to get a new copy of the workflow.
+
+
+      .. code-block:: bash
+
+         cylc get-resources tutorial/cylc-forecasting-workflow
+         cd ~/cylc-src/cylc-forecasting-workflow
+
+
+   #. **Break the ``post_process`` task**
+
+      Change the task script for the ``post_process`` task in
+      ``flow.cylc`` to make it fail:
+
+      .. code-block:: diff
+
+         - script = post-process $CYLC_TASK_PARAM_site 60
+         + script = exit 1
+
+   #. **Install, Play and inspect the workflow**
+
+      And check that ``post_process`` fails.
+
+      .. spoiler:: Step-by-step
+
+         .. code-block:: bash
+
+            cylc validate   # always good practice
+            cylc install
+            cylc play cylc-forecasting-workflow
+            # Either view in TUI or GUI, or run:
+            cylc cat-log cylc-forecasting-workflow
+            # Inspect job log for failed task:
+            cylc cat-log cylc-forecasting-workflow//<cycle point>/post_process
+
+
+.. nextslide::
+
+.. ifnotslides::
+
+   You can copy any changes in the :term:`source directory` to the
+   :term:`run directory` using:
+
+.. code-block::
+
+   cylc reinstall <workflow_id>
+
+.. ifnotslides::
+
+   But this does not reload the running workflow. To do that you need to use:
+
+.. code-block:: bash
+
+   cylc reload <workflow_id>
+
+
+.. ifnotslides::
+
+   .. note::
+
+      You may prefer to use ``cylc pause`` before reloading the workflow to
+      make if clearer which tasks were run before and after your changes.
+      After reinstalling and reloading the workflow use ``cylc play`` to
+      resume the workflow.
+
+
+.. practical::
+
+   .. rubric:: In this practical we will fix the workflow we broke in the
+      first practical, then re-run the broken task.
+
+   #. **Fix the ``post_process`` task**
+
+      Change the task script for the ``post_process`` task in
+      ``flow.cylc`` to make it pass again:
+
+      .. code-block:: diff
+
+         - script = exit 1
+         + script = post-process $CYLC_TASK_PARAM_site 60
+
+   #. **Use Cylc Reinstall and Reload to fix the workflow**
+
+      .. code-block:: bash
+
+         cylc reinstall cylc-forecasting-workflow
+         # You can do this in the cylc tui or GUI
+         cylc reload cylc-forecasting-workflow
+         cylc trigger cylc-forecasting-workflow//<cycle point>/post_process
+
+   #. **Check that the workflow has finished**
+
+      Have a look at the workflow log:
+
+      .. code-block::
+
+         cylc cat-log cylc-forecasting-workflow
+
+      You should see that ``post_process`` has succeeded:
+
+      .. code-block::
+
+         INFO - [20221027T1300Z/post_process_exeter running job:02 flows:1] => succeeded
+
+.. ifslides::
+
+   Next section: :ref:`Rose Tutorial <tutorial-rose-configurations>`