Console

The console reports the design-level results of a solve, the ones the per-element tables cannot show at a glance. It is organised as four pages, reached with the four buttons at the bottom of the panel, and the four share a footer giving the computation times and the convergence of the solver. The pages are filled when the solver runs, and they hold their contents until the next solve or until something invalidates them, such as a change of heat carrier fluid.

The console doubles as the help viewer: the help button of any panel replaces the results with that panel's help page, and any of the page buttons brings the results back.

Every solve refills all four pages and leaves the one you were reading in front. If a solve fails to converge, pages 1 to 3 stay empty, the status bar reads Convergence not reached and the faulty nodes are badged on the drawing; page 4 gives the reason.

Page 1, pumping and operating point - Reached with the button. The page opens with the borehole table described below, followed by one row per pump found in the drawing: the #ID of the pump node, its control mode, the flow Q in L/min, the head TDH in metres, the hydraulic power P_hyd and the electrical power P_elec in kW, and Cap., the share of the pump curve's capacity used at this operating point. Cap. turns bold from 90 %, the level at which the solver issues a warning, and shows a dash for a pump in ΔP-c or Q-c: in those two modes the setpoint is the pump, and no curve is consulted. When the drawing holds more than one pump, a Total row sums the two power columns. An asterisk on the electrical total means that at least one pump is missing from it because its curve carries no efficiency; hover it to see how many.

TDH is the head H the pump delivers at its operating point, and the hydraulic power is P = ρgQH at that point, with the density of the pump's own network at that network's temperature. The electrical power is P_hyd / (η_p·η_m), the pump and motor efficiencies coming from the Pump curve panel (Lamarche, 2023, Ch. 7); it shows a dash for a pump whose curve carries no efficiency. These are powers, not energies: the annual pumping energy also depends on the operating hours, which a hydraulic solve does not know.

The Ctrl. column gives the control mode with the acronyms used by circulator manufacturers, the same ones as in the Control list of the pump's Properties page. The setpoint of a ΔP-c or Q-c pump is read on that Properties page.

The Borehole U-loops table opens the page, since on a ground loop the first question is how much fluid reaches the least favoured borehole. A U-loop is recognised by its origin when the borehole was placed from the Ground heat exchanger family of the palette or generated by the Network Builder, whatever its orientation on the drawing. Otherwise it is recognised by its geometry: a 180° bend whose two legs rise nearly vertical and parallel. A U-bend with a third connection is never counted, since its two legs can no longer be told apart.

Three quantities are given for the boreholes of the field, each as a minimum, a mean and a maximum. Flow and Re are averaged over the two legs of a borehole. Loop Δp is summed over the two legs: it is the friction loss of the whole borehole, U-bend included, and therefore what the pump must overcome to push fluid through it. Like every friction figure of the console, it is R·Q², not the difference between the pressures at the two ends of a pipe; see ΔP is not the friction loss in the Hydraulic network chapter.

Flow alone does not tell whether the field works: a well-supplied borehole can still drop into the transition band, which degrades the heat exchange without showing in any flow figure. The Re column is there to catch it; page 2 gives the flow regime over the whole drawing.

The Ratio row divides the largest value by the smallest in each column. A ratio close to 1 means a balanced field. At 1.5, the most favoured borehole receives 50 % more fluid than the most starved one, and correct sizing elsewhere will not compensate for it. The extreme boreholes are named in the row labels, as in Minimum (#33), when the same borehole is the extreme of all three columns, which is the usual case on a field of identical boreholes. When the columns disagree, for instance on a field mixing bores or depths, the label reads Minimum alone.

Page 2, checks and balancing - Reached with the button. This page sums up the design checks of the Pipes table over the whole drawing. Two columns, Velocity and Friction, give for the same pipes the recommended range, the maximum reached and the pipe where it occurs, the number of pipes below and above the range, and the number of pipes checked (the last three rows are labelled Nb.). A non-zero count outside the range is set in bold. Both checks are advisory: they count the pipes outside the usual design window, 0.3 to 2.5 m/s and 100 to 400 Pa/m. The internal edges of composite devices are not pipes and are not counted.

The Friction column is a gradient, R·Q² per metre of pipe; a pressure difference would be dominated by ρgΔz and would flag every borehole leg. Below the range checks, Flow regime counts the pipes in each of the three regimes and gives the share of the head loss each regime carries. The share is the figure to read: thirty pipes in the transition band may be short connectors carrying almost no loss, while three pipes may carry most of it.

The transition row turns bold once it carries more than 1 % of the head loss. Between Re = 2300 and 4000, the friction factor is not given by a correlation but interpolated between the laminar and the turbulent values. The interpolated factor is lowest at Re = 2300 and highest at Re = 4000, so within that band a colder loop can show a smaller head loss and a higher flow than a warmer one. This is an effect of the interpolation, not of the flow physics.

Page 3, networks - Reached with the button. One column per independent network: the fluid actually used in the solve, the temperature and the density at which it was evaluated, the smallest and the largest pipe flow with their ratio, the critical head span, the delivered pump head, and the head-loss figures.

Min pipe flow, Max pipe flow and their Spread are counted over the pipes only, excluding the internal edges of devices and dead branches. The spread is a coarse indicator: a header naturally carries more than a branch, so a large ratio is expected and only says that the flows are not uniform.

Crit. ΔH is the difference between the highest and the lowest hydraulic head, P/ρg + z, among the nodes of the network. Its meaning depends on whether the network holds a pump.

On a network with no pump, such as a loop between two pressurization points or a branch fed from a header, the span is the head that the boundary conditions impose on the circuit, and it is the figure against which a pump can be sized before one has been drawn.

On a network with a pump, the span is not an independent check on the pump. Hydraulic head is a potential, so the span between the lowest and the highest head is the head the pump adds between its suction and its discharge, less the friction of the pipe that joins the pump to its neighbour. It therefore follows the reported TDH by construction and sits slightly below it, typically by 0.1 to 2 %. A TDH lower than the span cannot occur, and comparing the two will never reveal an undersized pump. A large gap between them would point to a numerical problem in the solve.

To judge whether a pump is right for its circuit, read the head it delivers against the loss the circuit demands: Total (kPa) on this page for the whole network, and the Loop Δp of the worst borehole on page 1.

The last block splits the head loss of each network in two: Friction loss, dissipated by the pipes along their length, and Minor loss, dissipated by the fittings through their K. Total, under a horizontal rule, is their sum and the only row of the page that adds up. Friction is R·Q² here, as on page 2. These figures are sums over all the pipes and fittings of the network.

Page 4, conservation and solver health - Reached with the button. Whether the solve converged, in how many Newton-Raphson iterations, the flow residual it reached, and the largest nodal flow imbalance with the node where it occurs. The imbalance is recomputed from the published pipe flows rather than taken from the solver, so it is an independent check: it should be several orders of magnitude below the smallest flow in the network.

Nodes at imposed pressure are left out of this check: an expansion vessel, a reservoir, a fill valve, or the anchor the solver places itself on a loop that declares none. Imposing a pressure removes the node's continuity equation, so the node may feed or drain the circuit. What it exchanges is reported on its own line, Boundary flow, with the node concerned. On a closed loop that flow is zero; on a circuit open between two pressurization points it is the flow crossing the circuit, which is a design figure. Solved at closes the table with the time of the solve.

Below a thin rule come the messages about the solve, in order of priority: Solver error, in red (a pump not wired in series, a circuit with two pressurization points, a check valve carrying reverse flow), then the networks left out of the computation, in orange, each with its reason on the line below.

The solver log follows under Details; this is where a convergence problem is explained. Not every entry is a fault: a network with no expansion vessel, reservoir or fill valve receives a pressure reference at its lowest-numbered node, at the static pressure of the Network panel, and the log says which node and at what pressure. Flows, velocities and head losses do not depend on that choice.

Computation time and convergence - The footer shared by every page reports the wall-clock time of each stage: Ass. builds the equations from the drawing, Solve runs the Newton-Raphson iteration itself, Post reorients the pipes and computes the aggregates, and Render repaints the scene and the tables. Total is their sum. Beside them, Convergence gives the iteration count and the flow residual of the solve.

On a network of a few dozen elements the whole cycle is a few milliseconds. On a field of several hundred boreholes the solve stays in the tens of milliseconds, and the rendering of the drawing becomes the dominant term.

GLN in practice: reporting a hydraulic study

The main result of a hydraulic study is the pump head, but a head on its own says little. Report it with the flow it corresponds to, the pump control mode, the fluid and its temperature, the maximum velocity and friction gradient with the number of pipes outside the recommended ranges, and the borehole ratio of page 1. All of these values are on the four console pages, and a reviewer can check the pump selection from them without opening the model.

Check the inputs first. The console only reflects the model and cannot tell whether a diameter, a roughness, a K or a length matches the project. Start with the lengths: a pipe marked with an orange bar in the Length column of the Pipes table still carries its seed length.

The velocity and friction ranges of page 2 are guidelines. A short connection above 2.5 m/s is often acceptable, whereas a field where every borehole leg runs below 0.3 m/s deserves a closer look, even though nothing on the page is flagged. The borehole ratio is more important: once the field is buried, an imbalance can only be corrected by throttling the favoured circuits, which adds pumping head for the life of the system.

Published benchmarks help judge whether the pumping power of a design is reasonable. Kavanaugh and Rafferty (2014, Ch. 6 §6.2, Impact of Pump Power) grade commercial ground-source systems by the power drawn by the pump motors per installed ton of cooling capacity, for a circulation of 2.5 to 3.0 gpm/ton (0.045 to 0.054 L/s per kW): grade A up to 50 W/ton (about 14 W per kW of cooling), B from 50 to 75, C from 75 to 100, D from 100 to 150 and F above 150. Page 1 gives the electrical power of each pump and their total, so the grade is simply that total divided by the installed cooling capacity. Converting a grade into a head requires the pump and motor efficiencies, which the table does not give.

Rhoda (2013) compares pipe-sizing criteria, piping configurations and pump control strategies on a simulated commercial building. Picard et al. (2017) derive the flow rate at which the extra pumping power offsets the gain in heat pump COP. For measuring and improving an existing pumping system, see U.S. DOE and Hydraulic Institute (2006).