diff --git a/xl2times/datatypes.py b/xl2times/datatypes.py
index 13f2d31..d26a9f2 100644
--- a/xl2times/datatypes.py
+++ b/xl2times/datatypes.py
@@ -207,6 +207,7 @@ def __init__(
             self.column_aliases,
             self.row_comment_chars,
             self.discard_if_empty,
+            self.query_columns,
             self.known_columns,
         ) = Config._read_veda_tags_info(veda_tags_file)
         self.veda_attr_defaults, self.attr_aliases = Config._read_veda_attr_defaults(
@@ -354,6 +355,7 @@ def _read_veda_tags_info(
         Dict[Tag, Dict[str, list]],
         Iterable[Tag],
         Dict[Tag, Set[str]],
+        Dict[Tag, Set[str]],
     ]:
         def to_tag(s: str) -> Tag:
             # The file stores the tag name in lowercase, and without the ~
@@ -374,6 +376,7 @@ def to_tag(s: str) -> Tag:
         valid_column_names = {}
         row_comment_chars = {}
         discard_if_empty = []
+        query_cols = defaultdict(set)
         known_cols = defaultdict(set)
 
         for tag_info in veda_tags_info:
@@ -395,6 +398,8 @@ def to_tag(s: str) -> Tag:
                     else:
                         field_name = valid_field["name"]
 
+                    if valid_field["query_field"]:
+                        query_cols[tag_name].add(field_name)
                     known_cols[tag_name].add(field_name)
 
                     for valid_field_name in valid_field_names:
@@ -411,10 +416,18 @@ def to_tag(s: str) -> Tag:
                     discard_if_empty.append(tag_name)
                 if base_tag in row_comment_chars:
                     row_comment_chars[tag_name] = row_comment_chars[base_tag]
+                if base_tag in query_cols:
+                    query_cols[tag_name] = query_cols[base_tag]
                 if base_tag in known_cols:
                     known_cols[tag_name] = known_cols[base_tag]
 
-        return valid_column_names, row_comment_chars, discard_if_empty, known_cols
+        return (
+            valid_column_names,
+            row_comment_chars,
+            discard_if_empty,
+            query_cols,
+            known_cols,
+        )
 
     @staticmethod
     def _read_veda_attr_defaults(
diff --git a/xl2times/transforms.py b/xl2times/transforms.py
index 901b826..837737b 100644
--- a/xl2times/transforms.py
+++ b/xl2times/transforms.py
@@ -3,7 +3,7 @@
 from pathlib import Path
 import pandas as pd
 from dataclasses import replace
-from typing import Dict, List
+from typing import Dict, List, Set
 from more_itertools import locate, one
 from itertools import groupby
 import re
@@ -17,17 +17,6 @@
 from . import datatypes
 from . import utils
 
-query_columns = {
-    "pset_set",
-    "pset_pn",
-    "pset_pd",
-    "pset_ci",
-    "pset_co",
-    "cset_set",
-    "cset_cn",
-    "cset_cd",
-}
-
 csets_ordered_for_pcg = ["DEM", "MAT", "NRG", "ENV", "FIN"]
 default_pcg_suffixes = [
     cset + io for cset in csets_ordered_for_pcg for io in ["I", "O"]
@@ -740,13 +729,16 @@ def fill_in_missing_values_table(table):
     return result
 
 
-def expand_rows(table: datatypes.EmbeddedXlTable) -> datatypes.EmbeddedXlTable:
+def expand_rows(
+    query_columns: Set[str], table: datatypes.EmbeddedXlTable
+) -> datatypes.EmbeddedXlTable:
     """
     Expand entries with commas into separate entries in the same column. Do this
     for all tables except transformation update tables.
 
-    :param table:       Table in EmbeddedXlTable format.
-    :return:            Table in EmbeddedXlTable format with expanded comma entries.
+    :param query_columns: List of query column names.
+    :param table:         Table in EmbeddedXlTable format.
+    :return:              Table in EmbeddedXlTable format with expanded comma entries.
     """
 
     def has_comma(s):
@@ -1720,6 +1712,7 @@ def is_year(col_name):
         if table.tag == datatypes.Tag.tfm_ins_ts:
             # ~TFM_INS-TS: Gather columns whose names are years into a single "Year" column:
             df = table.dataframe
+            query_columns = config.query_columns[datatypes.Tag(table.tag)]
             if "year" in df.columns:
                 raise ValueError(f"TFM_INS-TS table already has Year column: {table}")
             # TODO: can we remove this hacky shortcut? Or should it be also applied to the AT variant?
@@ -1807,7 +1800,9 @@ def process_transform_tables(
             df = table.dataframe.copy()
 
             # Standardize column names
-            known_columns = config.known_columns[table.tag] | query_columns
+            known_columns = (
+                config.known_columns[table.tag] | config.query_columns[table.tag]
+            )
 
             # Handle Regions:
             if set(df.columns).isdisjoint(
@@ -2032,6 +2027,7 @@ def make_str(df):
     if tag in tables:
         start_time = time.time()
         df = tables[tag]
+        query_columns = config.query_columns[tag]
         dictionary = generate_topology_dictionary(tables, model)
 
         df["process"] = df.apply(
@@ -2044,6 +2040,7 @@ def make_str(df):
         cols_to_drop = [col for col in df.columns if col in query_columns]
 
         df = expand_rows(
+            query_columns,
             datatypes.EmbeddedXlTable(
                 tag="",
                 uc_sets={},
@@ -2051,7 +2048,7 @@ def make_str(df):
                 range="",
                 filename="",
                 dataframe=df.drop(columns=cols_to_drop),
-            )
+            ),
         ).dataframe
 
         tables[tag] = df
@@ -2520,5 +2517,11 @@ def expand_rows_parallel(
     tables: List[datatypes.EmbeddedXlTable],
     model: datatypes.TimesModel,
 ) -> List[datatypes.EmbeddedXlTable]:
+    query_columns_lists = [
+        config.query_columns[datatypes.Tag(table.tag)]
+        if datatypes.Tag.has_tag(table.tag)
+        else set()
+        for table in tables
+    ]
     with ProcessPoolExecutor(max_workers) as executor:
-        return list(executor.map(expand_rows, tables))
+        return list(executor.map(expand_rows, query_columns_lists, tables))