diff --git a/Buildings/Occupants/BaseClasses/binaryVariableGeneration.mo b/Buildings/Occupants/BaseClasses/binaryVariableGeneration.mo
index e5b614fdacd..0e2f8fbc581 100644
--- a/Buildings/Occupants/BaseClasses/binaryVariableGeneration.mo
+++ b/Buildings/Occupants/BaseClasses/binaryVariableGeneration.mo
@@ -1,10 +1,11 @@
 within Buildings.Occupants.BaseClasses;
 function binaryVariableGeneration "Binary variables random generator"
+  extends Modelica.Icons.Function;
   input Real p(min=0, max=1) "Probaility of 1";
   input Integer globalSeed "Seed for the random number generator";
   output Boolean y "Random number";
 protected
-  Integer localSeed "Local seed";
+  Integer localSeed = 0 "Local seed";
   Integer state[Modelica.Math.Random.Generators.Xorshift1024star.nState];
   Real r(min=0, max=1) "Generated random number";
 algorithm
@@ -24,6 +25,11 @@ Higher <code>p</code> indicates a higher chance of generating <code>true</code>.
 </html>", revisions="<html>
 <ul>
 <li>
+October 3, 2023, by Michael Wetter:<br/>
+Initialized <code>localSeed</code>.<br/>
+This is for <a href=\"https://github.com/lbl-srg/modelica-buildings/issues/3549\">#3549</a>.
+</li>
+<li>
 July 20, 2018, by Zhe Wang:<br/>
 First implementation.
 </li>
diff --git a/Buildings/Occupants/BaseClasses/exponentialVariableGeneration.mo b/Buildings/Occupants/BaseClasses/exponentialVariableGeneration.mo
index 3cb4a5068a7..52b31fbb3d0 100644
--- a/Buildings/Occupants/BaseClasses/exponentialVariableGeneration.mo
+++ b/Buildings/Occupants/BaseClasses/exponentialVariableGeneration.mo
@@ -1,10 +1,11 @@
 within Buildings.Occupants.BaseClasses;
 function exponentialVariableGeneration "Random variable generator from the exponential distribution"
+  extends Modelica.Icons.Function;
     input Real mu  "Mean exponential distribution";
     input Integer globalSeed "Seed for the random number generator";
     output Real y "duration of event";
 protected
-    Integer localSeed;
+    Integer localSeed = 0 "Local seed";
     Integer state[Modelica.Math.Random.Generators.Xorshift1024star.nState];
     Real r "Generated random numberin the range 0 < random ≤ 1";
 algorithm
@@ -24,6 +25,11 @@ chance to generate a larger output <code>y</code>.
 </html>", revisions="<html>
 <ul>
 <li>
+October 3, 2023, by Michael Wetter:<br/>
+Initialized <code>localSeed</code>.<br/>
+This is for <a href=\"https://github.com/lbl-srg/modelica-buildings/issues/3549\">#3549</a>.
+</li>
+<li>
 July 20, 2018, by Zhe Wang:<br/>
 First implementation.
 </li>
diff --git a/Buildings/Occupants/BaseClasses/linear1D.mo b/Buildings/Occupants/BaseClasses/linear1D.mo
index 8e990ece457..243579a0d16 100644
--- a/Buildings/Occupants/BaseClasses/linear1D.mo
+++ b/Buildings/Occupants/BaseClasses/linear1D.mo
@@ -1,5 +1,7 @@
 within Buildings.Occupants.BaseClasses;
 function linear1D "Mapping a continuous input to a binary output through a linear relation"
+  extends Modelica.Icons.Function;
+
   input Real x "Continuous variable";
   input Real A=1 "Slope of the linear function";
   input Real B=0 "Intercept of the linear function";
diff --git a/Buildings/Occupants/BaseClasses/logit1D.mo b/Buildings/Occupants/BaseClasses/logit1D.mo
index 2f4b00eeaff..6c6ce03f270 100644
--- a/Buildings/Occupants/BaseClasses/logit1D.mo
+++ b/Buildings/Occupants/BaseClasses/logit1D.mo
@@ -1,5 +1,7 @@
 within Buildings.Occupants.BaseClasses;
 function logit1D "Mapping a continuous input to a binary output through a logistic relation"
+  extends Modelica.Icons.Function;
+
   input Real x "Continuous variable";
   input Real A=1.0 "Logistic relation: Slope";
   input Real B=1.0 "Logistic relation: Intercept";
diff --git a/Buildings/Occupants/BaseClasses/logit1DQuadratic.mo b/Buildings/Occupants/BaseClasses/logit1DQuadratic.mo
index 48ddb78773b..9cfd91c5069 100644
--- a/Buildings/Occupants/BaseClasses/logit1DQuadratic.mo
+++ b/Buildings/Occupants/BaseClasses/logit1DQuadratic.mo
@@ -1,5 +1,7 @@
 within Buildings.Occupants.BaseClasses;
 function logit1DQuadratic "Mapping a continuous input to a binary output through a quadratic logistic relation"
+  extends Modelica.Icons.Function;
+
   input Real x "Continous variable";
   input Real A=1.0 "Parameter defining the quadratic logistic relation";
   input Real B=1.0 "Parameter defining the quadratic logistic relation";
diff --git a/Buildings/Occupants/BaseClasses/logit2D.mo b/Buildings/Occupants/BaseClasses/logit2D.mo
index dc5e1e87889..5422d3ec2a7 100644
--- a/Buildings/Occupants/BaseClasses/logit2D.mo
+++ b/Buildings/Occupants/BaseClasses/logit2D.mo
@@ -1,5 +1,7 @@
 within Buildings.Occupants.BaseClasses;
 function logit2D "Mapping two continuous inputs to a binary output through a 2-dimension logistic relation"
+  extends Modelica.Icons.Function;
+
   input Real x1 "The first input variable";
   input Real x2 "The second input variable";
   input Real A=1.0 "Parameter defining the 2D logistic relation: mutiplier for the first input";
diff --git a/Buildings/Occupants/BaseClasses/weibull1DOFF.mo b/Buildings/Occupants/BaseClasses/weibull1DOFF.mo
index 837956c0982..9b8413d8a43 100644
--- a/Buildings/Occupants/BaseClasses/weibull1DOFF.mo
+++ b/Buildings/Occupants/BaseClasses/weibull1DOFF.mo
@@ -1,5 +1,7 @@
 within Buildings.Occupants.BaseClasses;
-function weibull1DOFF "Mapping a continuous input to a binary output through a Weibull Distribution Relation"
+function weibull1DOFF "Mapping a continuous input to a binary output through a Weibull distribution relation"
+  extends Modelica.Icons.Function;
+
     input Real x "Continous variable";
     input Real u=1.0 "Parameter defining the Weibull distribution threshold";
     input Real L=1.0 "Parameter defining the Weibull distribution normalization factor";
@@ -16,11 +18,11 @@ annotation (
 Documentation(info="<html>
 <p>
 This function generates a random binary variable with a continuous inputs
-<code>x</code> from a Weibull Distribution relation.
+<code>x</code> from a Weibull distribution relation.
 </p>
 <p>
 The probability of being 1 is calculated from the input <code>x</code> from a
-Weibull Distribution relation with three predefined parameters <code>u</code>
+Weibull distribution relation with three predefined parameters <code>u</code>
 (threshold, the output would be 0 if <code>x</code> is bigger than <code>u</code>),
 <code>L</code> (normalization faction) and <code>k</code> (shape factor).
 Then a random generator generates the output, which should be a binary variable.
diff --git a/Buildings/Occupants/BaseClasses/weibull1DON.mo b/Buildings/Occupants/BaseClasses/weibull1DON.mo
index 00e69d3c049..4fd5a94cedc 100644
--- a/Buildings/Occupants/BaseClasses/weibull1DON.mo
+++ b/Buildings/Occupants/BaseClasses/weibull1DON.mo
@@ -1,5 +1,7 @@
 within Buildings.Occupants.BaseClasses;
-function weibull1DON "Mapping a continuous input to a binary output through a Weibull Distribution Relation"
+function weibull1DON "Mapping a continuous input to a binary output through a Weibull distribution relation"
+  extends Modelica.Icons.Function;
+
     input Real x "Continous variable";
     input Real u=1.0 "Parameter defining the Weibull distribution threshold";
     input Real L=1.0 "Parameter defining the Weibull distribution normalization factor";
@@ -16,11 +18,11 @@ annotation (
 Documentation(info="<html>
 <p>
 This function generates a random binary variable with a continuous inputs
-<code>x</code> from a Weibull Distribution relation.
+<code>x</code> from a Weibull distribution relation.
 </p>
 <p>
 The probability of being 1 is calculated from the input <code>x</code> from a
-Weibull Distribution relation with three predefined parameters <code>u</code>
+Weibull distribution relation with three predefined parameters <code>u</code>
 (threshold, the output would be 0 if <code>x</code> is less than <code>u</code>),
 <code>L</code> (normalization faction) and <code>k</code> (shape factor). Then
 a random generator generates the output, which should be a binary variable.
diff --git a/Buildings/Occupants/BaseClasses/weibullVariableGeneration.mo b/Buildings/Occupants/BaseClasses/weibullVariableGeneration.mo
index 135a8c88a7a..f765a5f1abb 100644
--- a/Buildings/Occupants/BaseClasses/weibullVariableGeneration.mo
+++ b/Buildings/Occupants/BaseClasses/weibullVariableGeneration.mo
@@ -1,11 +1,13 @@
 within Buildings.Occupants.BaseClasses;
 function weibullVariableGeneration "Random variable generator from the Weibull distribution"
+  extends Modelica.Icons.Function;
+
     input Real lambda "Parameter defining the Weibull distribution scale factor";
     input Real k "Parameter defining the Weibull distribution shape factor";
     input Integer globalSeed "Seed for the random number generator";
     output Real y "Random variable generated from Weibull Distribution";
 protected
-    Integer localSeed;
+    Integer localSeed = 0 "Local seed";
     Integer state[Modelica.Math.Random.Generators.Xorshift1024star.nState];
     Real r "Generated random numberin the range 0 < random ≤ 1";
 algorithm
@@ -14,21 +16,26 @@ algorithm
     y := lambda*(Modelica.Math.log((1 - r)^(-1)))^(1/k);
   annotation (Documentation(info="<html>
 <p>
-This function generates a random variable, from a Weibull distribution with the 
-inputs of <code>lambda</code> and <code>k</code>. The random variable might be 
+This function generates a random variable, from a Weibull distribution with the
+inputs of <code>lambda</code> and <code>k</code>. The random variable might be
 the duration of a specific event, for instance the time to keep the HVAC on.
 </p>
 <p>
-The inputs <code>lambda</code> and <code>k</code> defines the probability density 
-function. <code>lambda</code> is similar to the mean value of exponential 
-distribution, and <code>k</code> defines the shape. A value of <code>k</code> = 1 
-means the Weibull distribution reduces to an exponential distribution. Genrally 
-speaking, higher <code>lambda</code> and higher <code>k</code> indicate a higher 
+The inputs <code>lambda</code> and <code>k</code> defines the probability density
+function. <code>lambda</code> is similar to the mean value of exponential
+distribution, and <code>k</code> defines the shape. A value of <code>k</code> = 1
+means the Weibull distribution reduces to an exponential distribution. Genrally
+speaking, higher <code>lambda</code> and higher <code>k</code> indicate a higher
 chance to generate a higher output.
 </p>
 </html>", revisions="<html>
 <ul>
 <li>
+October 3, 2023, by Michael Wetter:<br/>
+Initialized <code>localSeed</code>.<br/>
+This is for <a href=\"https://github.com/lbl-srg/modelica-buildings/issues/3549\">#3549</a>.
+</li>
+<li>
 July 20, 2018, by Zhe Wang:<br/>
 First implementation.
 </li>
diff --git a/Buildings/package.mo b/Buildings/package.mo
index 686419a1fa2..009f6dfc586 100644
--- a/Buildings/package.mo
+++ b/Buildings/package.mo
@@ -256,6 +256,17 @@ have been <b style=\"color:blue\">improved</b> in a
                        This is for <a href=\"https://github.com/ibpsa/modelica-ibpsa/issues/1798\">IBPSA, issue 1798</a>.
     </td>
 </tr>
+<tr><td colspan=\"2\"><b>Buildings.Occupants</b>
+    </td>
+</tr>
+<tr><td valign=\"top\">Buildings.Occupants.BaseClasses.binaryVariableGeneration<br/>
+                       Buildings.Occupants.BaseClasses.exponentialVariableGeneration<br/>
+                       Buildings.Occupants.BaseClasses.weibullVariableGeneration
+    </td>
+    <td valign=\"top\">Initialized <code>localSeed</code>.<br/>
+                       This is for <a href=\"https://github.com/lbl-srg/modelica-buildings/issues/3549\">#3549</a>.
+    </td>
+</tr>
 <tr><td colspan=\"2\"><b>xxx</b>
     </td>
 </tr>
