Low Delta-T Syndrome Diagnosis

Abriliam Consulting — Industrial Energy Management

Low delta-T syndrome is one of the most common and costly problems in chilled water systems. When the temperature difference between supply and return water drops below design, the system must push more water to deliver the same cooling — dramatically increasing pump energy. It also degrades chiller performance by reducing evaporator heat transfer effectiveness.

This notebook diagnoses low-delta-T conditions in the dataset using a combination of threshold analysis, scatter diagnostics, and efficiency segmentation.

Initial Scatter Diagnostics

Plant kW/ton vs Wet-Bulb Temperature: At low wet-bulb conditions (below ~14°C), the plant should operate very efficiently since the chiller has a low lift. Instead, we see a cluster of high kW/ton points — the plant is working harder than it should when conditions are favorable. This is a hallmark of low-delta-T syndrome.

Plant Tons vs Calculated Evaporator Tons: Comparing the load signal against the evaporator-side calculation (flow x delta-T) reveals measurement consistency. Deviations suggest either flow measurement error or genuinely degraded heat transfer.

=== Low-WB thresholds (computed within low-WB subset) ===
 n_points  WB_cut_C  bad_eff_thr_(top_q)  low_dt_thr_(bot_q)  high_flow_thr_(top_q)  %bad_eff  %low_dt  %high_flow  %classic_signature
      351      14.0                 5.72               4.128                 93.393    10.256   10.256      10.256                 0.0
=== Top 20 worst-efficiency hours within WB ≤ 14.0°C ===
  wb_C  plant_kw_per_ton   tons  chw_dT_C  chw_flow_m3h  chw_sup_C  chw_ret_C  chw_pump_kw  plant_kw   occ
11.035             7.591 40.000     6.557        19.288      6.411     12.968        6.000   303.639 0.353
12.349             7.416 40.000     5.835        21.103      6.471     12.306        6.000   296.654 0.169
11.275             7.370 40.000     6.299        18.926      6.417     12.716        6.154   294.781 0.064
11.468             7.203 40.000     3.911        33.551      6.594     10.505       12.791   288.102 0.308
11.033             7.106 41.937     6.172        21.577      6.191     12.363        6.576   297.999 0.295
 9.290             6.972 46.516     6.497        21.207      6.143     12.640        8.213   324.317 0.082
 9.132             6.942 40.000     5.853        21.457      6.155     12.008        6.000   277.672 0.110
10.329             6.881 46.003     4.772        30.524      6.224     10.995       10.476   316.536 0.309
11.216             6.866 56.495     3.216        53.598      6.505      9.721       26.659   387.872 0.195
10.008             6.802 40.000     6.043        20.280      6.226     12.269        8.941   272.066 0.165
 8.546             6.674 40.000     6.003        21.358      6.397     12.400       13.101   266.944 0.152
11.093             6.596 56.312     3.794        47.327      6.617     10.411       17.358   371.464 0.114
10.402             6.542 40.000     5.625        21.622      6.424     12.049        7.303   261.663 0.246
11.510             6.518 44.844     5.786        23.022      6.261     12.048        7.182   292.299 0.131
 7.674             6.509 40.000     6.240        19.443      6.371     12.611        6.000   260.350 0.147
 8.714             6.507 47.670     6.366        22.564      6.082     12.448        7.881   310.210 0.396
11.317             6.430 45.190     6.241        21.627      6.376     12.617        6.909   290.557 0.018
10.489             6.373 41.647     5.835        22.480      6.466     12.302        6.000   265.424 0.387
10.780             6.347 47.399     6.260        22.699      6.213     12.473        6.777   300.818 0.193
10.419             6.308 40.000     5.561        21.231      6.249     11.810        6.770   252.301 0.125

Classic signature count: 0 / 351 (WB ≤ 14.0°C)

Low Wet-Bulb Analysis Results

Isolating hours where wet-bulb temperature is below 14°C and examining the joint occurrence of:

The "classic signature" — all three conditions present simultaneously — identifies hours where the plant is definitively suffering from low-delta-T syndrome. The highlighted points in the kW/ton vs WB scatter plot cluster at low wet-bulb, confirming these are not weather-driven efficiency issues but hydraulic ones.

The delta-T vs flow diagnostic plot (colored by kW/ton) shows the expected pattern: as flow increases and delta-T drops, efficiency degrades. The dashed threshold lines partition the space into "healthy" and "syndrome" regions.

Power Floor Analysis

The Plant kW vs Tons scatter reveals a minimum power floor — even at very low cooling loads, the plant consumes a baseline amount of energy for pumps, fans, and ancillary systems. This power floor is particularly important because it means the plant's kW/ton metric degrades rapidly at part-load. Combined with low-delta-T syndrome (which increases pump power), part-load hours become disproportionately expensive.

Estimated plant power floor ≈ 142 kW
Incremental kW per ton ≈ 2.87 kW/ton
Inefficient hours (kW/ton > 4.0): 233 / 1344 (17.3%)
=== Summary by efficiency class ===
          n  wb_med  tons_med  plant_kw_med  dt_med  flow_med    pct
ineff                                                               
False  1111   20.33    231.20        812.63    5.58    133.29  82.66
True    233   11.90     95.96        422.10    5.43     53.67  17.34

Inefficiency Classification

Flagging hours where plant kW/ton exceeds 4.0 (a generous threshold) and examining their characteristics:

This analysis provides the evidence base for recommending operational changes: minimum load staging, CHW reset strategies, and pump speed optimization can all reduce the severity of low-delta-T syndrome.